Author: Jonas Neilson

  • Keyword research that starts with intent, not the volume column

    Keyword research that starts with intent, not the volume column

    You open a keyword tool. You sort by search volume. You start at the top.

    Two months later you’ve got a handful of posts chasing big, fuzzy terms, and almost none of them have brought in a single lead.

    It’s a really common trap, and it catches smart teams too. The problem isn’t effort. It’s that the volume column answers the wrong question first.

    So in this guide, I’ll show you the keyword research process I actually use. Start with the person and the job they’re trying to get done. Map the intent. Only then check demand. You’ll get a step-by-step workflow, free tools, a scoring table you can copy, and a look at why all this matters even more now that Google AI Overviews, AI Mode, and ChatGPT answer questions before anyone clicks.

    Why the volume column misleads you

    Search volume feels objective. It’s a number, and numbers feel safe.

    But volume only tells you how many times a phrase was typed. It doesn’t tell you who typed it, what they wanted, or whether your page could ever be the best answer.

    That’s a lot of missing information.

    Most searches live in the long tail

    Here’s the part most people miss. Ahrefs reports that keywords with fewer than 10 searches a month make up almost 93% of its U.S. keyword database.

    Ninety-three percent. Read that again.

    Google has also said for years that a big chunk of what people type is brand new. In its 2019 post on BERT, Google wrote that 15 percent of the queries it sees every day are ones it hasn’t seen before.

    So if you only work from what a tool can show you, you’re working from a partial map.

    Your tools hide the smallest, most specific queries

    Even Google’s own tool has blind spots. The Keyword Planner help page notes that keywords with very low search volume aren’t discoverable or forecastable.

    That’s fine for planning ads. For SEO, though, those “invisible” queries are often the ones sitting closest to a buying decision.

    AI answers change which keywords are worth a page

    Now for the new wrinkle.

    Pew Research Center analyzed the browsing data of 900 U.S. adults and found that Google users clicked a traditional result in 8% of visits when an AI summary appeared, versus 15% when one didn’t.

    The same study found AI summaries showed up for just 8% of one- or two-word searches, but for 53% of searches with 10 or more words. And 60% of queries starting with “who,” “what,” “when,” or “why” triggered one.

    Put simply: the long, question-shaped searches that intent-based research digs up are exactly where AI answers now live. You still want those topics. You just have to plan the page so it’s worth citing and worth clicking.

    What “search intent” really means

    Search intent is the job behind the search. What’s the person trying to get done when they type it?

    Most SEOs sort intent into four buckets. Here’s a quick reference you can keep open while you work.

    IntentWhat the searcher wantsExample queryPage that usually fits
    InformationalTo learn or understand somethinghow does schema markup workGuide, explainer, how-to
    CommercialTo compare options before buyingbest SEO audit tools for small sitesComparison, review, “best of” list
    TransactionalTo act or buy nowhire technical SEO consultantService page, product page, pricing
    NavigationalTo reach a specific brand or pageseopya contactHome, contact, or brand page

    Real searches often blend two of these. “Technical SEO audit cost” is part research, part shopping. That’s fine. What matters is that you decide which job your page will do best.

    My intent-first keyword research process

    Here’s the whole workflow, start to finish. It works for a five-page local business site, and it works for a big content program.

    Step 1: Write down the jobs your customers hire you for

    Before you open any tool, list the problems your best customers bring to you.

    Not keywords. Problems. Something like:

    • “Our traffic dropped after a Google update and we don’t know why.”
    • “We rank on Google, but we never show up in ChatGPT answers.”
    • “Our Amazon listings get views but don’t convert.”

    Pull these from sales calls, support tickets, onboarding forms, and the questions you get asked over and over. Shoot for 10 to 20 jobs.

    Step 2: Collect the questions people actually ask

    Now turn each job into the words people use. This is where you go wide.

    Good sources, most of them free:

    • Google’s own results. Search your seed term and write down the People Also Ask questions, related searches, and autocomplete suggestions.
    • Google Search Console. Your Performance report shows real searches that already bring up your pages.
    • Google Trends. See whether interest is rising, falling, or seasonal.
    • Question tools like AlsoAsked and AnswerThePublic, which show how questions branch off each other.
    • Communities. Reddit threads, Quora, industry Slack groups, and product reviews give you the exact phrasing people use when they’re frustrated.
    • Your own inbox. Real customer emails. Honestly, the most underrated keyword source there is.

    Step 3: Mine Search Console for long, conversational queries

    If your site already gets some traffic, this step is a gold mine.

    1. Open Search Console and go to Performance, then Search results.
    2. Set the date range to the last 3 months.
    3. Click + Add filter, choose Query, and switch the match type to Custom (regex).
    4. Paste a question pattern like ^(how|why|what|which|where|when|can|is|are|does|should) (the same pattern Ahrefs suggests in its long-tail guide).
    5. Sort by impressions and look for queries where you’re getting impressions but sitting past page one.

    Those are questions Google already connects with your site. A focused section or a dedicated page can often push them up.

    Step 4: Label every query with its intent

    Go back to your list and tag each query with one of the four intents from the table.

    The fastest, most honest way to do it? Look at the live results. Search the query and ask yourself:

    • Are the top results guides, comparison posts, product pages, or local listings?
    • Is there an AI Overview? A map pack? Videos? A shopping carousel?
    • Would the page you’re planning look like it belongs on this results page?

    If Google shows ten service pages and you were planning a 3,000-word guide, the results are telling you something. Listen to them.

    Step 5: Group queries into topics, not single keywords

    This is where most keyword research goes sideways. People build one page per phrase.

    Don’t.

    Instead, cluster queries that share the same intent and would be answered by the same page. “How long does SEO take,” “SEO timeline,” and “when will I see SEO results” belong on one page, not three.

    The quick rule I use: if the same pages rank for two queries, they probably belong to one topic.

    This matters for AI search too. Google says AI Mode uses a “query fan-out” technique that breaks a question into subtopics and runs many searches at once. A page that covers a whole topic cluster gives those sub-queries more to find on one URL.

    Step 6: Only now, check demand and difficulty

    Now you open the volume column. But you use it as a tiebreaker, not a starting point.

    For each topic cluster, add up the estimated volume of the queries in it. Then look at how tough the current results are. Huge brands? Or forums and thin pages you could beat?

    Tools like Ahrefs, Semrush, and Google’s Keyword Planner all give you estimates. Treat them like estimates. Different tools often disagree, so compare the trend and the rough size rather than chasing a precise number.

    Step 7: Score and prioritize your topics

    Last step. Score every topic so you know what to write first. Here’s the simple scoring table I use. Copy it into a spreadsheet.

    FactorQuestion to askScore 1 to 3
    Business fitDoes this topic lead naturally to what we sell?3 = directly, 1 = loosely
    Intent matchCan we build the type of page the results reward?3 = yes, 1 = not really
    ExperienceDo we have first-hand expertise or examples to add?3 = deep, 1 = none
    DifficultyCan we realistically compete with the current results?3 = yes, 1 = unlikely
    DemandIs there meaningful or growing interest?3 = strong, 1 = minimal

    Add up the scores and start with the highest totals. You’ll notice business fit and experience often outweigh raw demand.

    That’s the whole point.

    A worked example: an independent bike shop

    Let’s make this concrete with a made-up example. Say you run an independent bike repair shop.

    The volume-first approach would go after something like “bikes.” It’s huge, it’s vague, and it’s owned by national retailers and brand sites. Good luck.

    The intent-first approach starts with jobs:

    • “My gears keep slipping and I don’t know why.”
    • “I need a tune-up before a charity ride.”
    • “Is it worth fixing my old road bike or should I replace it?”

    Those turn into topic clusters like:

    1. Gear and shifting problems (informational): a troubleshooting guide on slipping gears, noisy chains, and when to bring the bike in.
    2. Bike tune-up (transactional, local): a service page with what’s included, pricing, turnaround time, and booking.
    3. Repair or replace (commercial): an honest guide with a simple decision checklist.

    Each cluster gets one strong page, and each page links to the next step. The guide links to the tune-up page. The tune-up page links to booking.

    None of these topics would win a volume sort. All three are a lot closer to revenue.

    How to turn your research into pages that win clicks and citations

    Great research is wasted if the page doesn’t match it. Here’s how I turn a topic cluster into an actual page.

    Answer the main question early

    Put a clear, direct answer to the core question in the first few paragraphs. Then go deeper.

    The reader who wants the short version gets it. And the page becomes easier for Google’s AI features and other answer engines to quote accurately.

    Use the real questions as subheads

    The questions you collected in Step 2 make great H2s and H3s. They mirror how people search and how AI systems break questions apart.

    Then make your title tag match the main intent too. My guide to writing title tags Google will actually use walks through that part.

    Add what only you can add

    Google’s guidance on helpful, people-first content asks whether content provides “original information, reporting, research, or analysis” and “insightful analysis or interesting information that is beyond the obvious.”

    That’s where your experience comes in. Real photos. Your process. The mistakes you see clients make. Specific numbers you can stand behind. I go deeper on this in E-E-A-T in practice.

    Connect the cluster with internal links

    Every page in a cluster should link to the others, and to the page where the reader takes the next step. That’s how topical depth becomes visible to people and crawlers.

    Want the full framework? Read internal linking as a system, not an afterthought.

    Make it easy for AI systems to cite you

    The intent-led pages you build here are also the ones most likely to get pulled into AI answers. Google’s documentation on AI features and your website says there are no extra technical requirements to appear in AI Overviews or AI Mode beyond being indexed and eligible for a snippet.

    So the fundamentals carry you most of the way: crawlable pages, clear answers, accurate facts, and a recognizable brand. My guide on how to rank better in LLMs and AI search covers the rest.

    Keyword research mistakes to avoid

    These are the shortcuts I see most often, and what I’d do instead.

    • Starting with the biggest number. Start with the job. Use volume to break ties.
    • One page per keyword. Build one strong page per topic cluster and intent.
    • Ignoring zero-volume queries. If customers ask it, it’s real demand, even when a tool says zero.
    • Guessing intent. Look at the live results before you write a word.
    • Treating tool numbers as facts. They’re estimates. Compare trends and rough size.
    • Researching once and never again. Revisit your list every quarter. Search behavior keeps shifting, especially with AI features in the mix.

    Your intent-first keyword research checklist

    Run through this before you green-light any new page:

    • We know which customer job this page serves.
    • We collected real questions from at least three sources.
    • We checked the live results and labeled the intent.
    • The page covers a topic cluster, not a single phrase.
    • We scored the topic on business fit, intent match, experience, difficulty, and demand.
    • The main question gets a direct answer near the top.
    • The page links to related posts and to a clear next step.

    How this shows up when we work together

    At Seopya, keyword research is never a spreadsheet we hand over and walk away from. We start from your customers’ jobs, map intent against the live results in Google and in AI answers, and build a content plan that ties back to the pages where you actually make money.

    If you’d like to see how that fits with technical work, content, and AI search visibility, take a look at our SEO and AI search services.

    Key takeaways

    • Search volume tells you how often something is typed, not why. Start with intent.
    • Most searches live in the long tail. Ahrefs puts keywords under 10 monthly searches at almost 93% of its U.S. database.
    • Long, question-style searches are where AI summaries show up most, according to Pew’s 2025 analysis. Plan those pages to be cited and clicked.
    • Cluster queries into topics and build one strong page per cluster.
    • Score topics on business fit, intent match, experience, difficulty, and demand, roughly in that order of importance.

    Frequently asked questions

    Is search volume still useful?

    Yes, as a tiebreaker. Use it to compare topics you’ve already decided fit your business and intent, and to spot rising or seasonal interest. Just don’t let it pick your topics for you.

    How many keywords should one page target?

    Think in topics, not counts. One page should go after one main intent and the cluster of closely related searches that share it. If two queries need different kinds of pages, split them.

    What free tools are best for intent-based keyword research?

    Start with Google itself (People Also Ask, related searches, autocomplete), Google Search Console, and Google Trends. Add a question tool like AlsoAsked, and read the community threads where your customers hang out.

    Does keyword research still matter with AI search?

    More than ever. AI answers lean on pages that answer specific questions clearly. Pew found that longer, question-style searches are far more likely to trigger an AI summary, and those are exactly the searches intent-first research uncovers.

    How often should I redo my keyword research?

    Review your priority topics every quarter, and check Search Console monthly for new queries. Big shifts, like a core update or a new AI feature, are a good excuse to look sooner.

    Where to start this week

    Keyword research isn’t a hunt for the biggest number. It’s finding the questions your customers are already asking, then building the page that answers them best.

    If you only do one thing this week, do Step 1. Grab a doc and write down ten problems your customers have brought you in the last few months. You’ll probably find your next three blog posts hiding in that list.

    If you’d like help building an intent-first content plan for your business, take a look at our SEO and AI search services or contact me and we’ll talk it through.

  • How to rank better in LLMs and AI search

    How to rank better in LLMs and AI search

    For about a decade, SEO meant one thing for most of us: getting found in a list of blue links.

    That work still matters. But here’s what changed. A lot of people now get an answer first, from Google AI Overviews or AI Mode, from ChatGPT, from Gemini, from Perplexity. Only after that do they decide which source to open, if they open one at all.

    So you can rank fine in classic search and still be missing from the answer. Frustrating, right?

    Let’s clear something up first. “Ranking” in a large language model isn’t a position you can buy. It’s closer to this question: when a system needs a source, is your page easy to find, easy to read, and specific enough to quote without twisting what you said?

    Nobody can guarantee you a citation. Anyone who says they can is selling something. What you can do is remove the reasons you’d get skipped.

    That’s what this post is about. I’ll cover entity clarity, writing that’s worth citing, honest structured data, brand mentions, crawler access for AI bots, and a format that works for people and machines.

    What “ranking” in LLMs can reasonably mean

    Google is the clearest official voice on its own AI features. Its documentation on AI features and your website says there are no extra technical requirements to show up in AI Overviews or AI Mode. No special schema either. A page just has to be indexed and eligible to show in Google Search with a snippet.

    The same document repeats the usual SEO basics:

    • Allow crawling.
    • Use internal links.
    • Put important information in text.
    • Keep structured data consistent with the visible page.
    • Keep your Business Profile or Merchant Center info current, if those apply.

    And inclusion still isn’t guaranteed.

    ChatGPT, Gemini (the app), and Perplexity are different animals. Depending on the mode and the question, they may pull live sources, lean on training data, or do both. What you can influence is more down to earth: Can the right crawler reach your page? Is the page specific and clearly about a named entity? Can your claims be quoted accurately? Is your brand referred to the same way around the web?

    If someone promises you a reliable “LLM rank,” don’t take it on faith. I wouldn’t.

    Make the entity obvious

    Systems retrieve and cite more reliably when they can tell who you are, what the page is about, and how it fits with the rest of your site. That’s traditional entity SEO. And it’s become more important, not less.

    Here’s what that looks like day to day. Use one consistent name for your organization, your authors, and the products or services you write about. Your homepage, About page, bylines, and contact info should all agree. If you work in a specific place, say so in plain words, not only in the footer. Put the subject in the title, the H1, and the first couple of sentences.

    Ambiguity costs you. If a system can’t tell whether “Mercury” is a company, a planet, or a metal, it’ll go find a less confusing source.

    Thin, interchangeable posts don’t travel well in AI search either. When I read a page that could’ve been written about any brand in the category, I always ask: what would a model even hold on to here? A page that describes an actual method or a first-hand observation gives it something real. That’s the page that gets pulled.

    Write so a system can cite you without distorting you

    Answer engines like passages they can lift without having to fix the meaning. That changes how you write. What you target is only half of it.

    • Lead with the answer. Open important sections with a direct answer, then add the conditions and detail. A heading that’s a real question, followed by a plain answer, is much easier to retrieve than a slogan.
    • Make claims concrete and bounded. “Meaningful SEO trust often takes six to eight months of persistent work” is something a reader can use. “We deliver explosive growth” isn’t.
    • Credit your sources. If a fact isn’t yours, link to the official or named source. If you can’t name a source, don’t make up a statistic. Ever.
    • Keep the text complete. Google’s AI features guidance says plainly that important content should be available as text. Don’t bury the answer in an image or a tab that never renders as HTML.
    • Date your updates. When the substance changes, update the page and say what changed.

    Bonus: the same habits help your classic snippets and title tags. Two birds.

    Structured data, used honestly

    Structured data doesn’t come with a secret AI-ranking bonus. Google says you don’t need special schema to show up in AI Overviews or AI Mode, and that any structured data you do use should match the visible text on the page.

    So use schema when it accurately describes something already on the page: an Organization, a Person, an Article, a LocalBusiness, a Product, or an FAQ people can actually see.

    Don’t mark up reviews you don’t have. Don’t credit authors who didn’t write the piece. Don’t add FAQ markup for questions that never appear on the page. Inaccurate markup isn’t a shortcut. It’s a trust problem waiting to happen.

    When your Organization name, logo, and official profiles line up in the markup and on the page, you’ve given systems a cleaner identity to attach mentions to. That’s useful. It still doesn’t replace being mentioned and linked as a real source.

    Brand mentions, and the sites that already cite you

    Classic SEO taught us to care about links. They still matter.

    But unlinked mentions and consistent naming help too. They help a model connect a claim to a brand, especially when those mentions show up on sources the system already treats as reliable.

    Can you fake this? Not usefully. A burst of identical bios on low-quality directories won’t make you the go-to source for a technical question.

    What helps, slowly, is the same work that has always built authority. Publish pages other practitioners would actually cite. Be accurate when someone quotes you. Show up in places your buyers already trust.

    If your public footprint is thin, start by making the official record consistent. Then write a few genuinely specific pages instead of a big stack of interchangeable ones.

    Give AI crawlers a clear, deliberate path

    Retrieval requires access. This is the most technical part of the whole job, and it’s the part I see sites get wrong by accident more than any other.

    OpenAI documents its crawlers separately. OAI-SearchBot is used to surface websites in ChatGPT’s search features. Sites that opt out won’t show up in ChatGPT search answers, though they can still appear as navigational links. GPTBot is the one used to crawl content that may go into training generative AI foundation models.

    Those two controls are independent. Keep that in mind.

    ChatGPT-User is a fetcher that kicks in when a user asks for something. It’s not an automatic crawler in the same sense. OpenAI also notes a robots.txt change can take about a day to be reflected, and firewalls can block crawlers even when robots.txt lets them in.

    PerplexityBot is used to surface and link websites in Perplexity results, and according to Perplexity, it isn’t used to crawl content for foundation models. Perplexity-User handles user-initiated fetches and generally ignores robots.txt, since a person asked for the page.

    Google’s Search AI features rely on ordinary Search eligibility. That means Googlebot has to be able to crawl and index the page. Google also documents other product tokens, including Google-Extended, in its crawler overview. Those tokens don’t replace being indexable in Search.

    Here’s the audit I run:

    1. Read your robots.txt. Make sure you haven’t disallowed OAI-SearchBot, PerplexityBot, or Googlebot on the URLs you want visible, unless that’s a deliberate policy.
    2. Check your CDN, WAF, and bot-challenge rules. A 403, a JavaScript challenge, or a login wall can shut out a crawler that robots.txt allows. When a provider publishes IP ranges, use those official lists, not a one-time sample from your logs.
    3. Compare what the crawler sees with what a user sees. The crawler should get the same main content, not a blank shell that only fills in after a client-side app loads.
    4. Decide training access separately from search access. And write the decision down.

    That third one catches more sites than you’d think.

    What not to do

    Don’t buy “submit your site to 50 LLMs” services. Don’t paste a hidden block of machine-only answers at the bottom of every page. There’s no evidence it helps, and it tends to create thin, duplicate content.

    Don’t confuse blocking GPTBot with blocking all visibility in ChatGPT. If you want to appear in ChatGPT search, the crawler that matters is OAI-SearchBot.

    Don’t invent statistics, fake expert quotes, or schema that doesn’t match the page.

    And don’t assume you need a special file or a “GEO score” for Google AI Overviews or AI Mode. Google has said the opposite.

    How this work shows up in a real engagement

    When I work on AI search visibility at Seopya, I start with the same foundation I’d use for classic SEO. Are the important pages indexable? Internally linked? Written with real specificity?

    Then I check crawler access, entity consistency, and whether the pages are actually citable. It all lives inside our SEO and AI search services.

    Here’s the blunt version: if the site can’t be crawled or the content is interchangeable, no amount of “LLM optimization” will make it a preferred source.

    Key takeaways

    • You can’t buy a reliable spot inside ChatGPT, Gemini, Perplexity, or Google AI Overviews. You can make your pages easier to retrieve and safer to quote.
    • For Google AI Overviews and AI Mode, the official requirement is ordinary Search eligibility. No special schema.
    • Be consistent about who you are, write bounded claims, keep structured data in line with visible text, and allow the crawlers that power search visibility.
    • Mentions and citations around the web still help, as long as they’re real.

    FAQ

    Is GEO different from SEO?

    Generative engine optimization is a handy name for the work of becoming a source in AI answers. It doesn’t replace technical SEO, on-page clarity, or earned authority. For its own AI features, Google treats SEO fundamentals as the relevant work.

    Should I block AI crawlers?

    That’s a policy decision, not a default. If you want to be eligible for ChatGPT search or Perplexity results, allow the search crawlers those companies document. Training crawlers can be a separate call. Make the decision on purpose, then check that your security tools actually honor it.

    Does structured data get me cited by ChatGPT or Perplexity?

    There’s no official structured-data requirement that guarantees a citation in those products. Accurate markup can make your entity easier to interpret, especially for Google. It won’t make up for a page a crawler can’t reach, or claims too vague to quote.

    The next step

    Pick five pages that should already be sources. Check each one: Is it indexable? Internally linked? Clear about the entity? Does it give a direct answer? Is the schema consistent (if you use any)? Can the crawlers you want to allow actually reach it?

    Fix those five before you publish ten more.

    If you’d rather have help, get in touch. I can tell you what an AI search pass on your site would involve.

  • Amazon SEO for sellers: how listings earn organic visibility

    Amazon SEO for sellers: how listings earn organic visibility

    Here’s something sellers forget. Amazon is a search engine that happens to have a checkout button.

    If a shopper can’t find your listing, nothing else on your detail page matters. Not your photos, not your price, not your reviews.

    The same goes when someone asks an AI shopping assistant whether your product fits what they need. The assistant can only work with what you actually wrote, photographed, and filled in as attributes. It can’t read your mind. (Neither can shoppers, for what it’s worth.)

    A lot of sellers treat Amazon SEO as a keyword exercise. Keywords matter, sure. But they’re only half the job. Conversion, completeness, and a reliable offer are part of relevance. They’re not a separate project you start after you “do SEO.”

    So I’m going to walk you through how I approach Amazon SEO for sellers: what people mean when they talk about the ranking system, how to research and place keywords (backend search terms included), how to write a listing that converts, and how to make that listing useful to Rufus.

    What “A10” actually refers to (and what Amazon documents)

    You’ll hear people talk about “A9” and “A10” like Amazon published a numbered algorithm with release notes.

    It didn’t. Those labels are industry shorthand for whatever the current ranking system is.

    What Amazon does document is more useful anyway. In its own guide to Amazon SEO, Amazon describes search optimization as helping shoppers find and choose your product. Research the terms they use. Write a clear title and bullets. Add backend search terms. Keep the listing complete. It also warns that keyword stuffing doesn’t work well and can hurt your ranking.

    When I talk to sellers, I use a simpler model than some mythical list of 200 factors:

    • Relevance. Does the listing clearly match the search? The product type, the defining attributes, the use case?
    • Performance. When shoppers see it, do they click, buy, and stay happy enough that returns and bad feedback don’t drag it down?
    • Readiness. Is it in stock, in the right category, and able to ship?

    Get those three right and you’re ahead of a surprising number of listings.

    Keyword research that starts with the shopper

    People on Amazon are usually close to buying. And they tend to search in the language of the product itself: type, size, material, count, compatibility, and the job they need it to do.

    Amazon’s own keyword research guidance starts in the same spot: the words customers actually type when they’re looking for products like yours.

    Here’s the sequence I use:

    • Type a bunch of plain-language searches into the Amazon search box and write down the suggestions.
    • Read the titles, bullets, and attributes on well-established competing listings. You’re looking for shared language, not phrases to copy word for word.
    • Dig through your own reviews, questions, and return comments. That’s where you’ll find the words buyers use after they’ve lived with the product for a while. It’s gold.
    • If you have Search Query Performance or Brand Analytics, look at which searches get impressions and convert, and which ones you show up for without earning the click or the sale. Those gaps are usually listing problems. Missing keywords are rarely the whole story.
    • Group your terms by intent: product type, specifying attributes, use cases, and synonyms.

    Then place them honestly.

    Your title carries the product type and the few attributes that set your offer apart. Your bullets explain benefits and specifics in normal sentences. Backend search terms are for relevant synonyms that didn’t fit anywhere visible.

    Backend search terms: the Generic keyword field

    Amazon lets you add keywords to your product’s catalog data that shoppers never see. According to Amazon, these backend search terms can still help a product rank for specific searches. Think of them as overflow, not a second stuffed title.

    Amazon’s current instructions, from that same Amazon SEO guide, are pretty simple:

    1. In Seller Central, go to Inventory, then Manage All Inventory.
    2. Edit the listing and open the Product details tab.
    3. Enter your terms in the Generic keyword field.

    Use lowercase letters. Separate words with spaces only. Include synonyms, abbreviations, and alternate names. Don’t repeat anything already in your title, bullets, or description, and don’t add terms that aren’t relevant. Skip articles and prepositions too, since they just waste space.

    I also keep a short “never” list, which lines up with Amazon’s own seller guidance: no competitor brand names, no ASINs, no punctuation as separators, and no claims that don’t describe the product.

    My rule of thumb? If a term would surprise a shopper who landed on your page, it doesn’t belong there. The field is limited. Put in your highest-value leftover terms and stop.

    Listing copy, media, and conversion

    A listing that ranks and then doesn’t convert won’t hold that rank for long.

    Shoppers (and Amazon’s systems) respond to whether the page lets someone make a confident, accurate decision. So here’s what I go through on every listing:

    • Title. Lead with the product type and the attributes that set it apart. Write it so a person could read it out loud. If you have to squint to find the actual product in a pile of keywords, the title’s doing the wrong job.
    • Bullets. Answer what a careful shopper wants to know before buying: what it is, what it fits, what’s included, how it’s used, and what’s genuinely different. Specifics beat adjectives every time.
    • Images and video. Show the product clearly, in use, and at a scale people can trust. Texture, contents, dimensions, what’s in the box.
    • Attributes and category. Fill in the structured fields: size, material, color, compatibility, count, all of it. Empty attributes hide you from filters and give AI assistants less to work with.
    • A+ Content. If you’re in Brand Registry, use it for comparisons, materials, and use cases that won’t fit in five bullets.
    • Offer quality. Price, stock, fulfillment, and a listing that matches what actually ships all feed into the experience Amazon is trying to protect. No amount of SEO copy fixes a product that’s always out of stock or keeps getting returned because the page overpromised.

    Reviews affect whether a shopper takes the risk on you. Earn them by delivering what the listing promised. Don’t try to manufacture them.

    And answer real Q&A clearly. Those answers become part of the public record for your product.

    Making the listing useful to Rufus (and Alexa for Shopping)

    In February 2024, Amazon introduced Rufus, a generative AI shopping assistant trained on Amazon’s catalog, customer reviews, community Q&As, and information from across the web. Shoppers can ask what to look for in a category, compare options, or ask questions right on a product page. Amazon later brought Rufus and Alexa together in the store as Alexa for Shopping.

    The name might change again. What it means for you won’t.

    An assistant can only cite, compare, and recommend based on information that exists. If your bullets never say who the product is for, what it works with, or how to use it, the assistant can’t make those answers up faithfully. Empty attributes give it even less to stand on.

    So what do you do about it? Honestly, the same things a good listing already needs:

    There’s no separate “Rufus SEO” campaign with a secret field. It’s listing quality, done thoroughly enough that a hurried shopper and a chatty assistant can both use the page.

    What not to do

    Don’t stuff your title with every synonym your keyword tool spat out. Amazon has said stuffing can hurt ranking, and it also lowers the click-through rate you need in the first place.

    Don’t put competitor brand names or ASINs in your backend search terms.

    Don’t create duplicate listings to chase slightly different keywords. You’ll split your reviews and give Amazon two weak pages instead of one strong one.

    And don’t treat Sponsored Products as a patch for a weak title, thin images, or a messy catalog. Ads can support organic visibility by building sales history and showing you which searches convert. They can’t cover for a listing shoppers don’t understand or trust.

    How this work shows up in a real engagement

    When I take on Amazon SEO, I start with the catalog, not with bids. Keyword research against real shopper language. A pass over titles, bullets, media, and attributes. A careful backend field. And a hard look at conversion and offer quality.

    Seopya led AirScent’s SEO, on Amazon and in search, through the brand’s consumer launch in December 2023 and the growth that followed. The public case study is on our Amazon SEO for sellers page. I’m mentioning it as context for how this work gets done, not as a promise that every catalog will move the same way.

    If you want help applying the same method to your products, our SEO and AI search services include Amazon SEO for sellers.

    Key takeaways

    • “A10” is industry shorthand, not a formula Amazon published. Plan around relevance, listing quality, and the customer’s experience.
    • Use shopper language honestly. Title and bullets first, backend terms only for what didn’t fit, and no competitor brands or repeats.
    • Conversion is part of Amazon SEO. Images, attributes, A+ Content, stock, and a listing that matches the product all count.
    • Write the listing so an AI shopping assistant can answer real questions from your content, not from guesswork.

    FAQ

    Does Amazon still use an “A10” algorithm?

    Amazon doesn’t publish an algorithm by that name. Sellers and agencies use “A10” as shorthand for the current ranking system. The safest planning assumption is the one Amazon’s own materials support: match the search, present a complete listing, and earn the sale.

    Are backend keywords more important than the title?

    No. Your visible listing is the main place to use shopper language, because it has to work for both the search system and the person looking at the result. Backend search terms are a supplement for synonyms you couldn’t fit in naturally. Overflow, not a hiding spot for stuffing.

    Can I optimize separately for Rufus or Alexa for Shopping?

    Not in the sense of a hidden flag or a special schema field. The assistant works from your listing content, attributes, reviews, and Q&A. The way to be more useful to it is to make the listing more complete and consistent. That’s the same work that helps a human shopper decide.

    The next step

    If your organic visibility is all over the place, start with one important ASIN and work through it in order. Research the searches. Rewrite the visible listing so a shopper can make a decision. Fill in the attributes. Clean up the backend field. Then (and only then) decide how ads should support it.

    Then move on to the next important product. One at a time beats all at once.

    If you’d rather have an experienced set of eyes on your catalog, get in touch. I’m happy to talk through what Amazon SEO would look like for your products.

  • Local SEO in the AI era: being found beyond the 3-pack

    Local SEO in the AI era: being found beyond the 3-pack

    Local search used to be pretty simple. A map pack, a few organic results underneath, done.

    Not anymore.

    People still search for a business “near me.” But now they also ask Google AI Overviews and AI Mode, ChatGPT, Gemini, and Perplexity which plumber to call, whether a restaurant is open late, or what they should know before walking in. And whatever answer they get is only as good as the public record those systems can reach about you.

    Here’s the good news. The fundamentals haven’t been replaced. NAP consistency, a complete Google Business Profile, and local pages that describe real services in a real place still matter. What’s changed is how many surfaces read that same record, and how much it costs you to be vague or inconsistent.

    So in this post, I’ll walk you through the local work that still has to be done well, and how to make it useful in a world of AI answers. We’ll cover NAP hygiene, your Business Profile, local pages and structured data, and the mentions that help a system tell your business apart from a similarly named one down the street.

    The local record is now read by more than the 3-pack

    Google has described local results for a long time as a mix of relevance, distance, and prominence.

    You can’t do much about distance. (Moving your shop isn’t an SEO tactic.) What you can control is whether your information is accurate, whether your categories and services match what people actually ask for, and whether the rest of the web refers to your business the same way.

    AI features sit right on top of that same record. Google’s documentation on AI features and your website says there are no extra technical requirements for AI Overviews or AI Mode. It also includes a very practical reminder: keep your Business Profile information up to date.

    A page still has to be indexed and eligible to show in Search with a snippet before it can appear as a supporting link. And for a lot of local queries, the Business Profile itself is part of what shoppers and systems are looking at.

    Other answer engines will pull from your website, your profile, directories, and reviews in ways you can’t fully see. So what’s the right response? It isn’t some special “local GEO” file. It’s making the official record consistent and your on-site pages specific enough to quote.

    NAP consistency, without the folklore

    NAP stands for name, address, and phone number.

    Consistency means the same public business name, the same storefront or service address, and the same phone number, written the same way, on your website, your Google Business Profile, and every listing you actually control.

    I’ll be honest. This is boring work. It’s also still the foundation, and I check it first on every local project.

    • Pick one public name and stick with it. If the sign on the door, your homepage, and your Business Profile all say slightly different things, you’re asking every system to guess. They won’t always guess right.
    • Put your full address and phone number in HTML text. On a contact or location page. Not only inside an image or a widget.
    • Use one address format. And if you’re a service-area business, don’t invent a street address. Follow Google’s rules on hiding or showing your address when you don’t serve customers at a storefront.
    • Update everything together. When you move or change your number, fix the site, the profile, and your listings in the same sitting.
    • Skip the bulk citation packages. A smaller set of accurate listings on real local and industry sites beats hundreds of messy, low-quality ones.

    Got more than one location? Give each one its own clear NAP and its own profile. And don’t mix addresses on the location pages.

    Google Business Profile: the details that still do the work

    For most local businesses, the Google Business Profile is the most important listing you have. It’s also one of the sources Google specifically tells you to keep current if you care about Search, AI features included.

    When I go through a profile, this is what I’m checking:

    • Primary and additional categories. Pick the category that best describes what you are. Add others only when they’re accurate. Categories aren’t a keyword field to cram things into.
    • Services and products. List what you really offer, in the words customers use. A half-empty list hides you from relevant searches and gives an assistant less to go on.
    • Hours, attributes, and special hours. Showing “closed” when you’re open (or forgetting holiday hours) costs you visits. It also chips away at trust.
    • Photos and description. Show the real place and the real work. Write a description that matches your website, not a paragraph of keywords.
    • Posts, Q&A, and reviews. Use posts and Q&A when you’ve got something real to say. Reply to genuine reviews. Don’t try to manufacture them.

    Google lays out its Business Profile policies in the Google Business Profile help center. The job here is accuracy and completeness. There’s no trick that “gets you into the 3-pack,” and pack placement isn’t something I can (or will) promise anyone.

    Local pages, structured data, and pages an AI system can cite

    Your Business Profile doesn’t replace a good local page on your own site.

    The page is where you explain who you serve, what you do, where you work, and what a customer should do next. It’s also the page you can make eligible for regular search results and for supporting links in AI features. You own it. Use it.

    A few things I look for:

    • Write for a real person in that area. Name the area naturally, describe your services in specific terms, and include your NAP in text. Please don’t spin up a “unique” city page for every nearby ZIP code if you’ve got nothing unique to say about it.
    • Answer the questions people ask before they call. What’s included, how appointments work, which areas you cover, and any limits they should know about.
    • Use real photos. Your actual location or your actual work, with honest alt text.
    • Keep the important stuff in visible text. If a system can’t read your hours, service area, or phone number from the HTML, it can’t cite them reliably.

    If you use structured data, be honest with it. LocalBusiness markup should match the visible name, address, phone number, hours, and URL. Don’t mark up reviews you don’t have. Don’t list a different address than the one on the page. Markup won’t make up for an incomplete profile or a thin, templated city page.

    One more thing. Your location page should be reachable from the main navigation or a locations hub, and from related service pages. A local page that’s only linked from the footer is easy to ignore, for people and for crawlers.

    Mentions, reviews, and being identifiable

    Local prominence has always been about more than your own site. Mentions in local publications. Membership in real associations. Sponsorships. Reviews from people who actually used you.

    In an AI-driven web, there’s an extra benefit: consistent naming across all those sources helps a model tell your business apart from another one with a similar name. If you’ve ever been confused with a competitor, you know why this matters.

    You can’t fake this with identical directory bios. What you can do:

    • Keep your official name consistent everywhere you control the listing.
    • Ask happy customers to leave a genuine review on the platform they already use, without incentives that break the platform’s rules.
    • Go after mentions that would make sense even if search didn’t exist. A local organization, a trade group, a real news item.
    • Make your About and contact pages match your profile, so a system checking more than one source finds the same business every time.

    It’s slow work. A new city-page template won’t replace it.

    What not to do

    Don’t build dozens of near-duplicate “city pages” that swap the place name and leave everything else identical. That isn’t local content. It’s also a weak source for any answer engine.

    Don’t stuff categories, reviews, or map embeds onto a page to look more local than you are. Don’t create a Business Profile for a location that doesn’t exist, or set up a service area in a way that breaks Google’s rules on how businesses can represent themselves.

    Don’t buy reviews or citations in bulk. Looking busy for a few months isn’t worth the long-term trust hit.

    And don’t assume you need a special schema or a “local AI” plugin to show up in AI Overviews. Google has said the opposite: keep your profile current and keep your site eligible for Search.

    How this work shows up in a real engagement

    When I take on local SEO at Seopya, I start with the record. That means the NAP on your website, your Business Profile, and the listings you already have.

    I fix the disagreements first. Then I improve the profile and the location pages so a person (and a retrieval system) can actually use them. All of this is part of our SEO and AI search services.

    What I won’t do is promise you a 3-pack spot or an AI Overview citation. What I will give you is a straight account of what’s inconsistent, what’s thin, and what needs to get more specific.

    Key takeaways

    • Local SEO still starts with an accurate, consistent public record: NAP, Business Profile, and specific local pages.
    • Google asks you to keep your Business Profile current, and local pages still need to be indexable to show up as supporting links in AI features.
    • Fill in categories, services, hours, and attributes accurately. They’re not keyword fields.
    • Use LocalBusiness structured data only when it matches what’s on the page.
    • Real reviews and genuine local mentions help. Bulk citations and cookie-cutter city pages don’t.

    FAQ

    Does NAP still matter if most people find me on Google Maps?

    Yes. Your profile, your site, and your listings should still agree with each other. Inconsistencies confuse customers, and they give every system a weaker record to work from.

    Can I optimize my local pages specifically for AI Overviews?

    Not with a special trick. Make your pages specific, accurate, and indexable. Keep the Business Profile current. Write answers a person would actually use. Google doesn’t describe any separate local optimization for AI Overviews or AI Mode.

    Should a service-area business still have a location page?

    Yes, if you’ve got something real to say about the service and the area you cover. Don’t invent a storefront address you don’t have, and don’t publish a separate page for every nearby town unless the content really is different. One clear service-area page plus an accurate profile usually beats twenty thin city pages.

    The next step

    Grab a notepad. Write down the exact public name, address, and phone number you want used everywhere. Now compare that to your homepage, your contact page, your Business Profile, and your top listings.

    Fix whatever doesn’t match. Fill in the profile. Then rewrite your local page so a customer (or an AI system) can actually use it.

    If you’d rather have help with that, get in touch. I’m happy to talk through what local SEO would look like for your business.

  • E-E-A-T in practice: experience, expertise, and being citable

    E-E-A-T in practice: experience, expertise, and being citable

    Here’s how a lot of people handle E-E-A-T. Add an author box. Mention “years of experience” somewhere. Cross your fingers that Google (and now the AI answer engines) decide the page is trustworthy.

    That’s not how it works.

    Experience, expertise, authoritativeness, and trust have to show up in the work itself. A nice bio can’t rescue a page that makes vague claims, hides who wrote it, or repeats what every other site already said.

    Google talks about this in its guidance on creating helpful, reliable, people-first content, and in more detail in the Search Quality Rater Guidelines, where E-E-A-T is a concept the raters use to judge pages. You can’t look it up in Search Console. There’s no special tag for it. It’s simply a way of asking: should this page be trusted for what it’s trying to do?

    And now there’s a second audience asking that question. ChatGPT, Gemini, Perplexity, and Google AI Overviews all need sources they can quote without twisting the meaning. A page that shows first-hand experience, names its author, and keeps its claims honest is a lot easier to cite than one that sounds confident and says nothing specific.

    So let’s get practical. I’ll show you how to put experience on the page, how to make expertise checkable, and how to write so a system can cite you accurately.

    What E-E-A-T is (and what it isn’t)

    The four letters stand for experience, expertise, authoritativeness, and trust.

    Trust is the big one. The other three are there to support it. A page can be written by a genuine expert and still fall flat if the claims are sloppy, the byline is missing, or the site around it looks careless.

    A few distinctions I’d keep straight:

    • It’s a quality concept, not a plugin metric. Google has said that content showing strong E-E-A-T is useful, especially on topics that can affect someone’s health, finances, or safety. That doesn’t mean any widget can tell you your “E-E-A-T score.” If one claims it can, be skeptical.
    • The extra E (experience), which Google added in late 2022, is about first-hand knowledge. You’ve used the product, done the work, visited the place, or actually been in the situation the page describes.
    • Authoritativeness is often earned off your site. Other reputable sources referring to you, consistently, for the thing you actually do.
    • Trust includes boring hygiene. A real identity, accurate contact details, clear policies where they’re needed, and pages that don’t contradict each other.

    If you remember only one line from this post, make it this one: E-E-A-T is demonstrated, not declared.

    Put experience where a reader can see it

    Experience is the easiest letter to fake in a sentence. It’s the hardest to fake across a whole page.

    “In my experience” followed by a generic tip isn’t experience. A specific observation is.

    On a service or advice page, real experience looks like:

    • The actual sequence you use, including the dull steps and the places where the method stops working.
    • Details someone rewriting a competitor’s article couldn’t have made up: the order you run an audit in, the one field in a form that actually matters, the mistake you see over and over and how you tell it apart from a false alarm.
    • Named, checkable context when you have it. And silence when you don’t. If you can’t publish a metric, don’t invent one. Point to a public case only when the facts are already public, like we do on the Amazon SEO page.

    On a product or “best of” page, experience means evidence that you used the thing: what you tested, under what conditions, and what you wouldn’t recommend. On a local page, it means knowing the place and the service area, beyond a city name dropped into a template.

    My test is simple. Strip out the brand name and the adjectives. Is anything specific left?

    If not, the page hasn’t shown experience yet.

    Make expertise and authorship checkable

    Expertise isn’t a job title in italics. It’s the reader being able to tell who’s responsible for the page, and why that person is in a position to say what they’re saying.

    When I review a site for this, here’s the pass I make:

    • Real bylines. The author should be a person (or a clearly identified team), not a ghost brand account. The exception is when the organization really is the author.
    • An about page that agrees with everything else. Publish an author or about page with the same name, the same role, the same location if that matters, and a short account of the work. Not a résumé stuffed with keywords.
    • Show your homework. When you rely on an official rule, link to the official document. When you’re offering a judgment call, say so, and say what it’s based on.
    • Keep it current. Update the page when the substance changes, and show the date. An expert who never revises a technical article eventually stops looking like one.
    • No phantom reviewers. Don’t attach an “expert reviewer” to a page that person never reviewed. If someone checked the work, say what they checked. If nobody did, don’t make one up.

    The same standard applies to the site around the article. A careful post on a site with no contact info, two different business names, and copy-pasted policy pages will still feel shady.

    Identity is part of expertise.

    Write claims that can be cited without distortion

    This is where old-school E-E-A-T and AI search overlap.

    Answer engines like sentences they can lift without having to fix the meaning. Funny enough, that’s exactly what a careful human reader wants too.

    How to write that way:

    • Say the point plainly, then add the conditions. “A page must be indexed and eligible for a snippet to appear as a supporting link in Google AI Overviews or AI Mode” is citable. It’s bounded, and in that case it’s documented. “Do this to win AI Overviews” is not.
    • Skip the implied guarantees. I won’t promise first-page rankings, citations, or lead volume. I’d be wary of anyone who does.
    • Separate facts, methods, and opinions. A reader (or a model) should be able to tell which is which.
    • Keep important answers in visible text. Not only in an image, a podcast, or a PDF.
    • Match your structured data to the page. If you use schema for an author, article, or organization, it has to reflect what’s visible. Schema that inflates credentials is a trust problem, not a trick.

    In my experience, a citable page is usually just a slightly more careful version of a useful page. It isn’t a different genre of content.

    What not to do

    Don’t buy author bios, fake credentials, or “expert” quotes. Don’t generate a thin article and then dress it up with a stock photo and a made-up reviewer.

    Don’t confuse volume with authority. Ten interchangeable posts on the same topic, each one announcing its own expertise, are weaker than one specific piece a practitioner would actually bookmark.

    Don’t treat an E-E-A-T checklist as the work itself. An author box, a headshot, and a “last updated” stamp are useful when they’re true. They don’t replace first-hand detail, accurate claims, and a site that stands behind its contact information.

    And don’t hide who runs the site. That goes double if you write about money, health, or legal issues. If readers can’t tell who’s responsible, why should they trust the advice?

    How this work shows up in a real engagement

    When I work on E-E-A-T at Seopya, I start with the pages that already have a job: service pages, key articles, the About page. Then I ask whether a skeptical reader could tell who wrote them, what first-hand work they’re built on, and which claims can be checked.

    From there I fix the identity gaps and tighten the writing. Only after that do I worry about how those pages connect to the rest of the site. It all sits inside our SEO and AI search services, because trust isn’t a separate campaign. It’s the quality of the source you’re asking Google and the answer engines to use.

    One more honest note. This kind of trust usually takes persistent work over months, not a weekend of bio updates. I say the same thing about SEO timelines in general, and it holds here too.

    Key takeaways

    • E-E-A-T shows up in the work, not in an author box. It isn’t a score in Search Console.
    • Experience means specific, first-hand detail, not “in my experience” stuck in front of a generic tip.
    • Authorship should be checkable, consistent across the site, and honest about who actually reviewed the page.
    • Bounded claims, linked sources, and answers in plain visible text make you easier to cite.
    • Never invent credentials, metrics, or reviewers.

    FAQ

    Is E-E-A-T a Google ranking factor?

    Google doesn’t present E-E-A-T as a single number you can tune on its own. It’s a concept for thinking about whether content is helpful and reliable, including in the Quality Rater Guidelines.

    Improving the substance, authorship, and accuracy of a page can support visibility. Buying the look of authority won’t.

    Does E-E-A-T help with ChatGPT or Perplexity citations?

    There’s no official “E-E-A-T tag” those systems look for. But a page that’s specific, attributable, and consistent is easier for an answer engine to use as a source than one that’s vague or anonymous. The overlap is real. It’s just not a guarantee.

    Do I need a personal byline on every page?

    No. Not every page needs a personal essay, and some organizational pages are rightly authored by the company. Someone accountable should still be easy to identify, though. And article-style advice is usually stronger with a real person attached, as long as that person actually did the work.

    The next step

    Pick three important pages and read them like a skeptic would.

    Can you tell who’s responsible? Is there any first-hand detail a competitor couldn’t have copied? Are the claims bounded, and when they lean on someone else’s rule, do they link to it?

    Fix those three pages before you roll out a new “trust” template across the whole site.

    If you’d like help with that pass, get in touch. I’m happy to talk through what would really need to change on your pages, and what a byline alone won’t fix.

  • Internal linking as a system, not an afterthought

    Internal linking as a system, not an afterthought

    Most sites I look at don’t have a keyword problem. They have a navigation problem.

    Important pages sit two or three clicks farther from the homepage than they should. Related articles never mention each other. Category hubs are thin. The only way a crawler (or a brand-new reader) finds a useful piece is through the sitemap or the search box. Then the team publishes more, and wonders why none of it builds on itself.

    Sound familiar?

    Internal linking is how you tell people and search engines which pages matter, how your topics connect, and which URL is the main source for a subject. Even Google’s guidance on AI features lists making content easy to find through internal links as one of the fundamentals that still apply. Not because links are trendy. Because a page nobody can discover can’t be indexed, ranked, or cited.

    So in this post I’m going to treat internal linking as a planned system. You’ll see which pages deserve the most links, how I write anchors, how to keep crawl paths short, and how to keep the whole thing from rotting as your site grows.

    What an internal linking system is for

    A good internal linking system does three jobs at once.

    It gives crawlers a path. Google finds pages by following links. When I run a site audit, I almost always find “orphan” URLs that exist in the sitemap or in analytics but aren’t linked from a single indexable page. If you want a page in the index, someone already on your site should be able to click their way to it.

    It tells everyone what a page means. When several related pages link to one URL with descriptive anchors, you’re saying, “this is the page we mean when we talk about this topic.” That helps traditional rankings. It also helps retrieval systems that want one clear, specific source instead of five posts that all sort of say the same thing.

    It helps the reader finish what they came to do. Someone who showed up for a definition might need a process next. Or a local page. Or your service page. Those paths should be written into the content on purpose, not left to a “related posts” widget that’s basically guessing.

    Here’s my honest take. If your site has been publishing for months with little to show for it, look at your links before you look at your content calendar. A lot of the time the missing work is already on the site. It’s just hard to reach. (I come back to this in what a real site health audit covers.)

    Start with a short map, not a plugin

    You don’t need a 400-row spreadsheet to get going. You need a map of the pages that should be collecting authority, and an honest look at what points to them right now.

    Here’s how I build that map:

    • Pick your 10 to 20 workhorse URLs. Service pages, key location pages, pillar articles, and the product or category pages that need to be found. These should get the most thought-through internal links on the site.
    • Check what already points to each one. Note which pages link to it, whether it’s in the main navigation, and which obviously related articles don’t mention it yet.
    • Crawl the site. Flag orphans, pages you can only reach from the footer, and paths that take four or more hops from the homepage to a money page.
    • Group content by topic, not by publish date. A post about title tags should point to a post about meta descriptions, and both should point to the on-page or services page that’s the main offer. “Previous post / next post” links based on dates aren’t a system.
    • Pick one main URL per topic. If two articles are competing for the same job, merge them. Or make the relationship obvious: one is the overview, the other is a narrower follow-up, and the links say so.

    Navigation and breadcrumbs matter. They’re just not enough on their own. Your main nav can only hold so many items before it stops being useful.

    The links that usually move things? The ones inside the body of relevant pages, sitting right where a reader would actually want to click.

    Write anchors that describe the destination

    “Click here.” “Read more.” A naked URL.

    All three waste the link.

    Your anchor text is a tiny description of where you’re sending someone. “SEO and AI search services” tells a person and a crawler a lot more than “learn more.” “How to write title tags that Google will actually use” beats “this post” every time.

    The rules I follow:

    • Use the words a knowledgeable reader would use to describe the destination. Not some stuffed variation of a keyword you’re trying to force in.
    • Link from a sentence that already makes a real point. The link should feel like the natural next step, not decoration.
    • One clear link to the right URL beats three repetitive links to the same page in one article.
    • Read the paragraph out loud. If the anchor sounds weird, it’s forced. Rewrite it.

    Keep crawl paths short, and keep important pages indexable

    Links only help if they can be followed.

    A page behind a login, blocked in robots.txt, tagged noindex, or buried inside a JavaScript interface that never renders links in the HTML isn’t really part of your system. It’s just sitting there.

    These are the patterns I watch for:

    • Orphans. Give every orphan at least one editorial link from a related, indexable page, and add it to the right hub or category.
    • Footer-only links. A footer link can help discovery. But a page that matters should also get links from contextual content, and from the navigation if it belongs there.
    • Facets and parameters. Ecommerce and directory sites can spin up thousands of filtered URLs. Link to the canonical, useful versions. Don’t let every sort order and checkbox combo turn into an internal destination.
    • Redirects and chains. Point internal links at the final 200 URL, not an old slug that hops twice before landing. It’s basic hygiene, and it keeps you from leaking whatever meaning the link was supposed to carry.
    • Pagination and archives. Category pages and date archives can make good hubs if someone actually curates them. They shouldn’t be the only road to your best work.

    People sometimes ask me for the official “maximum clicks from the homepage” rule. I don’t think there’s one worth treating as law. My test is simpler: if someone who already knows your site still can’t find the page without using search, the path is too long or too buried.

    Maintain the system as you publish

    Internal linking breaks down the moment you treat it as a one-time project.

    Every new page should get links from the pages already in its topic group. And it should link out to the hub or service page it supports. If you can’t name those connections when you hit publish, I’d ask whether the new page actually has a job.

    For most sites, a light routine is plenty:

    • When you publish, add two or three contextual links from older, related pages. Not a sitewide blast.
    • Once a quarter, recrawl your important URLs. Look for new orphans, broken internal links, and hubs that have gone stale.
    • When you update or merge a page, fix the anchors that pointed to the old version in the same sitting. Not three months later.
    • Keep an eye on automated “related article” modules. They can help, but they usually link by tag or recency, not by meaning. Use them as a backup, and check they aren’t pointing at thin or off-topic pages.

    None of this is glamorous.

    It’s also how a site starts to feel like a body of knowledge instead of a pile of posts.

    What not to do

    Don’t install a plugin that stuffs the same exact-match anchor into every post in a category. It looks like manipulation, and it’s almost always worse than fewer, better links.

    Don’t build an “HTML sitemap” page as a stand-in for editorial links and then leave the rest of the site disconnected. A list of URLs isn’t a reading path.

    Don’t nofollow your own internal links to try to “sculpt” PageRank. You’re far more likely to confuse the crawl than outsmart it.

    And please don’t link every mention of a word to the same sales page no matter the context. Readers notice. It makes the whole site feel cheaper.

    How this work shows up in a real engagement

    When I work on internal linking at Seopya, it’s usually part of the technical and on-page work, not a standalone “link install.” After the crawl, I map the hubs, fix orphans and chains, and add contextual links where they genuinely help a reader get where they’re going.

    That all sits inside our SEO and AI search services. A site people can’t navigate is a site that’s hard to rank or cite with any confidence.

    You should be able to see the result. Important URLs are easy to reach. Related pages acknowledge each other. And a new article has a home on the day it goes live.

    Key takeaways

    • Internal links are how people and crawlers find your pages, and how your site signals which URL owns a topic.
    • Start with a short list of important pages, then add links from related content. Navigation alone won’t cut it.
    • Write anchors in plain language that describe where the link goes.
    • Short paths, final 200 URLs, no orphaned money pages.
    • Every time you publish, merge, or update, touch the links too.

    FAQ

    How many internal links should a page have?

    There’s no magic number. Link when there’s a real next step for the reader or a real relationship between topics. A long article can carry more links than a short one. A short page stuffed with links is usually dodging clearer writing.

    Do links in the footer or sidebar count?

    They can help with discovery, especially for utility pages. But they’re weaker than a contextual link in the body of a related article, and they shouldn’t be the only links to a page you care about.

    Does internal linking help with AI Overviews or LLM citations?

    Indirectly, yes. Google includes internal linking among the fundamentals that still support AI features in Search, because a page has to be discoverable and indexable to be eligible at all.

    Will internal links “insert” you into an AI answer on their own? No. Nothing does that.

    The next step

    Want a simple place to start? Pick ten important URLs. Crawl to see how they’re reached today. Then add links to them from the related pages that should’ve been pointing there all along. Fix any broken or redirected internal links on those paths while you’re at it.

    That’s an afternoon of work, and it often does more than another blog post would.

    If you’d rather have help building the system on your site, get in touch. I’m glad to walk through what a crawl would show and which links would actually be worth adding.

  • What a real SEO site health audit actually covers

    What a real SEO site health audit actually covers

    Here’s a pattern I see all the time. A business publishes a blog post every week for a year. They pick up a few links. They tweak their homepage copy twice. And traffic just… sits there.

    Usually the problem isn’t the content.

    It’s something underneath it. Pages Google never indexed. A redirect chain that bleeds every internal link. A template that crawls along on phones. Three versions of the same URL quietly fighting each other. You can publish on top of that for a long time and never see anything compound.

    That’s why every serious project at Seopya starts with a site health audit. I don’t mean a one-click “score out of 100.” I also don’t mean a fifty-page PDF of warnings that nobody opens twice. I mean a structured look at whether search engines can find, crawl, render, and trust the site you already have.

    And yes, this matters for AI search now too. Google’s documentation on AI features says a page has to be indexed and eligible for a snippet before it can show up as a supporting link in AI Overviews or AI Mode. Other answer engines can only cite pages they can actually reach and read.

    So let me walk you through what a real audit covers, in the order I actually do it, and how to turn the findings into a short list of fixes instead of a pile of noise.

    What a site health audit is for

    A site health audit answers one question: what is stopping good pages from being found and understood?

    Everything else is secondary.

    Google breaks its own process into three stages (crawling, indexing, and serving results) in its guide to how Search works. I follow that same path in an audit, because a problem early in the chain caps everything after it. A beautifully written page that can’t be crawled won’t rank. It won’t get quoted in an AI answer either. It’s invisible, no matter how good the writing is.

    When I’m done, you should walk away with three things:

    • A clear picture of which pages search engines can reach, which ones they index, and which ones they ignore.
    • A short, prioritized list of technical problems, ranked by how much they hurt important pages (not by how many URLs a tool happened to flag).
    • A baseline you can measure the next six to eight months of work against.

    That’s it. If an audit doesn’t give you those, it was a report, not an audit.

    Step 1: Crawl the site the way a search engine would

    I always start with a full crawl. Screaming Frog, Sitebulb, whatever you like. The important part is setting it up to behave like Googlebot Smartphone.

    Why smartphone? Because Google uses mobile-first indexing, so the mobile version of your page is the one that counts. If your mobile template hides content, links, or structured data that the desktop version shows, that gap is a finding all by itself. You’d be surprised how often it shows up.

    What I look for in the crawl

    • Status codes. Broken internal links (4xx), server errors (5xx), and redirect chains or loops. Every internal link should go straight to a final 200 URL. No detours.
    • Orphan pages. Pages that sit in the sitemap or analytics but that no internal link reaches. If you don’t link to a page, you’re telling crawlers it doesn’t matter. They tend to believe you.
    • Crawl depth. How many clicks it takes to get from the homepage to your important service and product pages. Money pages buried five clicks deep are common, and they’re one of the easier things to fix.
    • Rendering. Whether key content and links actually exist in the rendered HTML. This one bites hardest on JavaScript-heavy themes and page builders.

    Robots.txt and crawl directives

    Next, I open robots.txt and read it line by line. It’s a short file. One wrong Disallow can still hide an entire section of your site.

    Keep one distinction in mind, which Google spells out in its robots.txt introduction: robots.txt controls crawling, not indexing. If you want a page out of the index, you use a noindex directive, as covered in Google’s guidance on blocking indexing. And the page has to stay crawlable so Google can actually see that directive.

    Blocking a URL in robots.txt and adding noindex to it at the same time? That’s one of the most common contradictions I run into. Google can’t read the noindex on a page it isn’t allowed to fetch.

    Step 2: Check what’s actually indexed

    A crawl shows you what exists. Google Search Console shows you what Google decided to do with it.

    The Page indexing report lists your indexed pages and sorts the rest by reason: “Crawled – currently not indexed,” “Discovered – currently not indexed,” “Duplicate without user-selected canonical,” “Excluded by ‘noindex’ tag,” and so on.

    Don’t panic when you see a big excluded number.

    The skill here is reading the report, not reacting to it. Plenty of URLs are excluded on purpose: filtered pages, tag archives, old parameters. What you care about is whether any important page landed in the wrong bucket. So I export the report, line it up against the crawl, and flag every page that should be indexed but isn’t.

    Canonicals and duplicates

    Duplicate URLs split your signals. The usual suspects:

    • HTTP and HTTPS versions
    • www and non-www
    • Trailing slash and no trailing slash
    • Uppercase paths, tracking parameters, and print or pagination versions

    Google’s documentation on consolidating duplicate URLs lays out your options: redirects, rel="canonical", and consistent internal linking.

    In the audit, I check that every important page has a self-referencing canonical, that canonicals point to indexable 200 URLs, and that the canonical Google picked (you can see it in the URL Inspection tool) matches the one you declared. When Google disagrees with you, it’s worth finding out why.

    XML sitemaps

    Your sitemap should be a clean list of the URLs you want indexed. Nothing redirected. Nothing noindexed. Nothing returning an error. Google’s sitemaps overview covers the format and the limits.

    Here’s a trick I use on every audit. I compare three lists: the sitemap, the crawl, and the indexing report. Where they disagree is usually where the problem lives.

    Step 3: On-page hygiene across templates

    This isn’t copywriting yet. That comes later.

    At this stage I’m doing a hygiene check across templates, because one broken template can repeat the same mistake on hundreds of pages. Fix the template and you fix them all.

    • Titles and meta descriptions: missing, duplicated, truncated, or identical across a whole section.
    • Headings: one clear H1 per page that matches what the page is for, with a sensible H2/H3 outline underneath.
    • Thin and near-duplicate pages: location or service pages that differ by a single word. (You know the ones.)
    • Images: missing ALT text, huge files, and filenames like IMG_4021.jpg.
    • Structured data: whether the markup is valid, matches what’s visible on the page, and describes the organization, author, and page type accurately. Google’s introduction to structured data is clear that markup should reflect what’s actually on the page.

    That last one is where old-school SEO and AI-era SEO meet.

    Clear titles, honest headings, consistent business details, and accurate markup help Google understand a page. That same clarity makes it easier for answer engines to pull out a correct answer and credit the right source. Same work, two payoffs.

    Step 4: Performance and page experience

    Speed isn’t a ranking trick. Think of it as a floor you need to stand on.

    Google reports Core Web Vitals on three metrics:

    • Largest Contentful Paint (LCP) for loading
    • Interaction to Next Paint (INP) for responsiveness
    • Cumulative Layout Shift (CLS) for visual stability

    INP replaced First Input Delay as a Core Web Vital in March 2024, as the Chrome team announced on web.dev. The published “good” thresholds are an LCP of 2.5 seconds or less, an INP of 200 milliseconds or less, and a CLS of 0.1 or less, measured at the 75th percentile of real visits.

    When I audit performance, I start with field data: the Core Web Vitals report in Search Console and Chrome UX Report data. That’s what real visitors experience. Then I use lab tools like Lighthouse to figure out why a template is slow. It’s usually some mix of giant hero images, render-blocking scripts, bloated page-builder output, and third-party tags nobody remembers adding.

    Google’s own page on Core Web Vitals and Search treats these metrics as part of page experience. I think that’s the right way to see them. They matter for users. They don’t replace relevance.

    Step 5: Turn findings into a priority list

    This is where most audits fall apart.

    A crawler will happily hand you thousands of “issues.” Your job is to figure out which ten actually matter. I score every finding on two questions:

    How many important pages does it affect? And how directly does it block crawling, indexing, or understanding?

    Then I sort them into this order:

    1. Blockers first: anything that stops important pages from being crawled or indexed. Robots.txt mistakes, stray noindex tags, broken canonicals, server errors.
    2. Signal leaks next: redirect chains, duplicate URL versions, orphaned money pages, and internal links pointing at redirects.
    3. Template fixes: one change that cleans up hundreds of titles, headings, or image attributes at once.
    4. Performance: the slowest templates with the most traffic, fixed at the source.
    5. Polish: the long tail of minor warnings, handled as normal maintenance.

    Every item gets an owner, a rough estimate, and a way to confirm it’s fixed. Usually that’s a re-crawl or a validation request in Search Console.

    That list is the deliverable. Not the raw export.

    What not to do

    A few traps I’d steer you away from:

    • Don’t chase a tool’s health score. A 98 out of 100 can hide a noindexed service page. A 70 can belong to a site that’s doing just fine.
    • Don’t start with content or outreach. New pages and new links just pour into the same leaks. Fix the plumbing first.
    • Don’t “fix” everything at once. If you change URLs, redirects, and templates in one big release, you’ll never know what helped and what hurt.
    • Don’t treat the audit as one and done. Sites change every time a plugin updates or someone adds a page. A lighter health check every quarter catches regressions before they get expensive.

    How this shows up in a real engagement

    At Seopya, the audit is the first phase of our SEO and AI search services. It happens before we write a word of new content or send a single outreach email.

    You get the prioritized fix list. We work through the blockers together. Only then do we move into content strategy, link building, and AI search visibility, because all of that depends on a site search engines can read cleanly. Building on a cracked foundation just makes a bigger crack.

    FAQ

    How long does a site health audit take?

    It depends on how big and complicated the site is. A small brochure site can be reviewed in days. A large ecommerce site with faceted navigation and international versions takes longer. Honestly, speed matters less than whether the findings get prioritized and verified.

    Can I run an audit myself with free tools?

    Yes, a basic one. Google Search Console, PageSpeed Insights, and the free tier of a desktop crawler will surface most blockers on a small site.

    The hard part is interpretation. Knowing which warnings are intentional and which ones are actually costing you visibility takes practice.

    Does technical SEO matter for AI search and AI Overviews?

    It does. Google’s AI features documentation says that to be eligible as a supporting link in AI Overviews or AI Mode, a page has to be indexed and eligible to show in Search with a snippet. There aren’t separate technical requirements. Other answer engines also need to reach and parse a page before they can cite it.

    Good technical health won’t guarantee a citation. Technical problems can absolutely rule one out, though.

    The next step

    If you’ve been publishing for months and don’t have much to show for it, don’t publish faster. Look underneath.

    Crawl the site. Compare it to what Google has indexed. Fix the blockers on your most important pages. Then build on top.

    And if you’d like an experienced set of eyes on it, take a look at our SEO and AI search services or get in touch. I’m happy to walk you through what an audit of your site would involve, and what it can and can’t tell you.

  • How to write meta descriptions that earn clicks

    How to write meta descriptions that earn clicks

    Let me guess. At some point, someone handed you a checklist that said meta descriptions must be exactly 155 characters, stuffed with your keyword, or Google would punish you.

    Forget most of that.

    A meta description is still one of the most useful bits of writing you’ll do for a page. Just not for the reasons those 2019 checklists claimed. It isn’t a secret ranking lever. It isn’t a hard 160-character field. And Google won’t always show the sentence you wrote.

    So what is it? A pitch. A short, accurate summary you offer up so that, when Google does use it, the right person has a reason to click.

    Updated November 2026.

    In this post I’ll show you how meta descriptions actually work now, how I write them so they earn the click, and how they fit alongside the on-page writing that AI Overviews, AI Mode, and other answer engines really read. For title tags (and how titles and descriptions should work as a pair), check out the companion guide, How to write title tags that Google will actually use.

    What a meta description is, and what it isn’t

    A meta description is an HTML element in your page’s head. Visitors don’t see it on the page. Search engines can use it as the snippet, that short blurb under the title link in the results.

    Google’s documentation on how snippets are created is what I point clients to. It’s a lot more precise than most blog posts on the subject:

    • Snippets are automatically created from page content, and they’re designed to preview the part of the page that best matches a specific search.
    • That means Google can (and often does) show different snippets for different queries.
    • Google may use your meta description when it thinks the tag describes the page better than an extract from the body.

    So you aren’t programming the snippet. You’re handing Google a well-written default, and making sure the page itself has clear sentences that can step in if Google rewrites it.

    Both jobs matter. A description that promises something the page doesn’t deliver will get ignored. Or worse, it’ll earn a click that bounces right back.

    That same document is clear on length, too. There’s no official character limit. Google truncates the snippet as needed, usually to fit the device width. The old “stay under 160 characters or Google will punish you” rule was never Google’s rule. And it’s even less useful now that most clicks happen on a phone.

    I still draft to a rough range of 140 to 160 characters, with the point of the sentence inside the first 120, because a phone only has so much width. But if a longer description is more accurate, I’ll take accurate and accept that the last clause might get cut. What I won’t do is pad a weak sentence with synonyms, or chop a good one at some arbitrary 155-character wall.

    Write unique descriptions that match search intent

    Google’s first best practice is one I still see broken on otherwise solid sites: create unique descriptions for each page.

    One site-wide blurb, or a template that just swaps the city name, doesn’t help anybody tell two pages apart. It also gives Google less reason to trust the tag.

    Unique means the description is true of this page:

    • A service page should name the service, who it’s for, and a reason to click.
    • A location page should be specific to that place, not a find-and-replace of your homepage.
    • A blog post should say what the reader will learn.
    • A product page can pull together facts scattered around the page (what it is, who it’s for, a distinguishing detail). Google actually encourages that use.

    Next, match the intent behind the searches you actually want. If people search “how to write meta descriptions,” they want a method, not a sales pitch for an agency. If they search “Miami SEO agency,” they want to know who you help and how. Write that answer in plain language. Google’s snippet guidance says outright that long strings of keywords don’t give users a clear idea of the page and are less likely to be shown.

    What about your brand name? It helps recognition. It shouldn’t eat the first line. I usually put the brand at the end, or leave it out if the title link already carries it. Lead with the benefit or the answer.

    Writing for SERP click-through rate

    A description is a tiny ad. An honest one.

    The goal is to earn a click from someone who’ll be glad they showed up. Not to trick a bigger crowd onto a page that won’t help them.

    When I write one, I run through four quick questions:

    1. Does the first clause answer the search?
    2. Is there a specific detail a competitor’s snippet probably doesn’t have?
    3. Would I still stand behind this sentence if Google showed it for a different search on the same page?
    4. Does it still make sense if a phone cuts off the last eight words?

    Specifics beat adjectives. “Expert, industry-dominating, results-driven SEO” tells a skeptic absolutely nothing. “A Miami SEO studio for brands that need to rank in Google and be cited in AI answers” tells them who it’s for.

    The title link does most of the heavy lifting. Your description should close the gap, not repeat the title word for word.

    Examples: before and after

    Quick note: these are illustrative rewrites, not results from a test.

    Service page, before (stuffed and vague): “Miami SEO, SEO Miami, best SEO company Miami, SEO services, search engine optimization, AI SEO, affordable SEO agency.”

    Ouch.

    After: “Technical SEO, on-page work, content, and local search for Miami brands, plus the foundations that help pages get cited in Google AI Overviews, ChatGPT, Gemini, Perplexity, and Grok.”

    Blog post, before (too short and generic): “Learn about meta descriptions and SEO.”

    After: “Google may rewrite your snippet, and AI Overviews need a page that can show one. Here is how to write unique meta descriptions that match intent, hold up on mobile, and still earn the click.”

    See the difference? The better versions name the real offer, drop the keyword list, and leave a clear takeaway. They’d still be useful if Google ignored the tag and grabbed a sentence from the opening paragraph instead. That’s exactly why I hold the on-page intro to the same standard.

    Mobile display, and why the opening sentence matters twice

    Most of your audience sees a shorter snippet than you do on your desktop monitor. Since truncation follows device width, the words that matter need to come early. Put the qualifier, the location, or the call to action at the end, not the start.

    Here’s the part people skip.

    The opening of the page matters just as much as the meta tag. When Google builds the snippet from the body, it often reaches for a short sentence near the top. An answer-first intro (one or two sentences that state the point before the background) gives Google, and anyone who lands on the page, the same clear start.

    Writing for AI Overviews, AI Mode, and other answer engines

    Classic SEO and AI search aren’t two separate crafts here. Google’s documentation on AI features and your website says there are no extra technical requirements and no special schema you need to show up in AI Overviews or AI Mode. To be eligible as a supporting link, a page has to be indexed and eligible to be shown in Google Search with a snippet.

    That last bit is easy to miss.

    A nosnippet robots directive, or a max-snippet of 0, doesn’t only hide your description in the blue links. Google’s robots meta tag documentation says those controls also stop your content from being used as a direct input for AI Overviews and AI Mode. So if you want to be eligible for those features, don’t switch snippets off. Google’s note on succeeding in AI search makes the same point: tighter preview permissions limit how your content can be featured.

    What about ChatGPT search, Perplexity, Gemini, and Grok? None of them publish a Google-style guide that says “we display your meta description.” I’m not going to pretend they do.

    In practice, answer engines cite pages they can fetch and understand. And the on-page habits that help Google pick a good snippet are the same ones that make a page easier to quote accurately:

    • A title, H1, and opening paragraph that agree with each other and name the entity clearly.
    • An answer-first introduction.
    • Headings that map to real questions.
    • Structured data that matches the visible text, as Google asks in its introduction to structured data.

    Open Graph title and description tags keep your public summary consistent when someone shares a link. Google also lists og:title as one of the sources for title links.

    Think of your meta description as the short public abstract for the page. If it’s unique, accurate, and lines up with the first screen of the page, you’ve given Google, an LLM, and a person pasting your URL into a chat the same honest summary.

    How to test descriptions in Search Console

    You won’t get a tidy lab result saying one sentence raised CTR by some exact percentage. Not unless you run a careful test over time.

    What you can do (and what I do after a round of rewrites) is watch the page in Google Search Console’s Performance report.

    Google’s guide to debugging Search traffic drops is the framework I lean on. If impressions hold steady but clicks drop, your title and snippet may not be helping people understand the page, or a competitor may have a more appealing result. After you update descriptions and titles, compare a period after the recrawl with a similar period before. Do it by URL and by query. The site-wide average hides too much.

    A few habits that keep me honest:

    • Write down the date you changed the copy.
    • Give Google time to recrawl before judging anything.
    • Watch position alongside CTR. If CTR shifts at the same time as a big jump in average position, it’s probably about rank, not your writing.
    • If you have enough data, check mobile and desktop separately.

    Don’t cloak, and don’t run an open-ended experiment that shows Google one title and users another. Google’s guidance on website testing is the standard I follow.

    One last thing. CTR is just clicks divided by impressions. It’s a useful diagnostic, not a trophy. A higher CTR on the wrong search is still a bad outcome if those visitors leave right away.

    A practical checklist

    When I review a page’s description, this is my list:

    • The page has its own meta description. Not shared with another URL, and not a lightly edited site tagline.
    • It accurately sums up this page, in the same language as the content, and matches the intent of the searches the page should win.
    • No keyword stuffing. The first clause still makes sense if a phone cuts off the rest.
    • The title, the H1, and the opening paragraph tell the same story. The description adds a reason to click. It doesn’t contradict them.
    • The facts a searcher wants (who it’s for, where, what they get) are in the sentence, not implied.
    • There’s no nosnippet or max-snippet:0 on a page I want eligible for AI Overviews or AI Mode.
    • Open Graph and social descriptions match the search description.
    • After publishing, I watch the URL in Search Console (queries, impressions, and CTR) once it’s been recrawled.

    FAQ

    Do meta descriptions still matter if Google rewrites them so often?

    Yes. Google says it may use your description when the tag is the more accurate summary, and a strong description is still the best default you can give it. Even when Google pulls the snippet from the page, a clear, unique summary and an answer-first intro are what it has to work with.

    Is there a character or pixel limit I must hit?

    Not officially. Google truncates the snippet to fit the device. Get the meaning in early, and treat roughly 140 to 160 characters as a planning range, not a rule.

    Are meta descriptions a ranking factor?

    Google doesn’t describe them that way. They affect whether someone understands your result and clicks. Useful, unique descriptions can improve the quality of the traffic you get. They don’t stand in for a page that deserves to rank.

    Should I write a different description for AI search than for Google?

    No. Write one accurate description, then make sure the intro, headings, and title agree with it. Eligibility for Google’s AI features depends on being indexed and snippet-eligible, not on some second hidden summary.

    The next step

    If you only do one thing after reading this, open your highest-impression pages and read their snippets like a stranger scrolling on a phone. Rewrite the ones that are vague, duplicated, or stuffed. Then line up the title, the H1, and the first paragraph, so the page would still earn the click even if Google ignored your tag completely.

    Titles deserve the same care, and the companion guide linked above covers them. If you’d like this done across your whole site, take a look at Seopya’s SEO and AI search services or get in touch.

  • How to write title tags that Google will actually use

    How to write title tags that Google will actually use

    You spend twenty minutes crafting the perfect title tag. You publish. A week later you search for your page and Google is showing something else entirely.

    Annoying? Sure. But it’s not a bug.

    The title tag is the most visible line of SEO writing you’ll ever do, and it’s also the one Google is most willing to swap out. Google’s systems build a title link (the clickable headline in results) from several sources. If your title is weak, stuffed, or out of date, Google will happily reach past it for something that describes the page better.

    Updated November 2026.

    In this post I’ll show you how title tags work now, how to write ones Google is more likely to actually use, and how titles and meta descriptions should be written as a pair. For the description half, see the companion guide, How to write meta descriptions that earn clicks.

    What a title tag is, and what a title link is

    The title tag is the HTML <title> element in your page’s head. It shows up in browser tabs, bookmarks, and a lot of shares.

    The title link is what Google actually shows in Search.

    They’re often the same text. They’re not guaranteed to be.

    Google’s documentation on influencing title links is the doc I keep open whenever I audit a site. It says title-link generation is automated, that it considers the page and references to it around the web, and that Google draws from a list of sources, including:

    • The <title> element
    • The main visual title on the page
    • Heading elements, such as the H1
    • The og:title meta tag
    • Other large, prominent text
    • Other text on the page
    • Anchor text, including links that point to the page
    • WebSite structured data

    Your official <title> is still the source Google uses most. But it’s one vote, not a command. If the tag is empty, boilerplate, inaccurate, or stale, Google may swap in your H1 or some other line.

    And after you change a title, Google has to recrawl and reprocess the page before the title link can update. That can take days or weeks. Patience helps.

    This is exactly why I write the title tag, the H1, and the opening paragraph together. When those three agree, you give Google fewer reasons to invent a headline. You also give a reader (or an answer engine) one consistent name for the page.

    Best practices that still hold

    Google’s own best practices are short. And honestly, most title problems I find come from skipping these basics.

    Give every page a descriptive, concise title. Skip vague labels like “Home” or “Blog.” There’s no official limit on how long a title element can be, but the title link gets truncated to fit the screen. So the useful part has to show up early enough to survive a phone.

    Don’t stuff keywords. A few descriptive terms help. Repeating the same phrase in slightly different forms doesn’t. Google notes that this kind of stuffing can make a result look spammy. (It usually does.)

    Don’t repeat boilerplate. Every page needs its own text. A catalog where every page is titled “Cheap products for sale,” or a long template that changes one word, makes results impossible to tell apart.

    Brand the title concisely. On interior pages, put your site name at the start or the end, separated by a hyphen, a colon, or a pipe. Keep it short. Your homepage can say a bit more about the site. Google may also drop the site name from the title link when it’s already shown next to the result.

    Make the main visual title obvious. One clear H1, visually bigger than everything else. That’s the simple fix. If two headings have the same size and weight, Google may just pick the first one.

    Write the title in the same language and script as the page. A mismatch gives Google a reason to replace your tag.

    The same document also lists the usual suspects behind rewrites: half-empty template titles, old years, inaccurate tags, micro-boilerplate, and no clear main title. If you see a title link that doesn’t match your tag, check that list first. Nine times out of ten, the answer’s there.

    How I write a title that earns the click

    I start with search intent, not a keyword list. What’s the person typing? And what would a useful headline say back to them?

    A good title does three things. It names the topic in the words people actually use. It adds a qualifier that sets this page apart, like a method, an audience, a place, or a year. And it leaves room for the brand at the end when the brand helps.

    I draft to a rough range of 50 to 60 characters. That’s because of truncation, not because Google scores length. If a precise title runs a little long, I’d rather be precise and put the unique words first. I don’t pad a short, clear title just to “fill the pixels.”

    Your title and H1 can be close without being identical. The title tag can carry the brand and slightly tighter wording for the results page. The H1 can be a bit more conversational.

    What they can’t do is disagree.

    A title that says “pricing” and an H1 that says “features.” A title that says 2024 on a page about 2026. Those are exactly the mismatches Google looks for.

    Then I read the first two sentences of the page. If the title promises a guide, the intro should start delivering on that promise right away. That’s good traditional SEO. It’s also how you write for systems that summarize a page from the top down.

    How titles and descriptions work together

    A search result is a tiny unit of copy: title link, URL, snippet.

    The title says what the page is. The description says why it’s worth opening. If they repeat each other, you’ve wasted the second line. If they contradict each other, you’ve told Google two stories and given the searcher a reason to hesitate.

    So I write them as a pair. The title carries the noun and the qualifier: “How to write title tags that Google will actually use.” The description finishes the thought: “Google builds title links from your title tag, your H1, and other sources, and it will rewrite a weak one. Here is how to write unique titles that match intent and work with your descriptions.”

    Keep both unique to the page.

    And remember, Google may replace either line. Your on-page H1 and your answer-first intro are your backup title and backup description. There’s a full method for the second line in the companion guide on meta descriptions.

    Examples: before and after

    These are illustrative rewrites, not measured test results.

    Homepage, before (vague, stuffed): “SEO, SEO Services, Best SEO Company, Search Engine Optimization, AI SEO | Seopya SEO SEO”

    After: “Expert SEO and AI Search Visibility | Seopya”

    Service page, before (boilerplate plus keyword pile): “Miami SEO | SEO Miami | SEO Agency Miami | Affordable SEO Services Miami Florida”

    After: “SEO and AI Search Services in Miami | Seopya”

    Article, before (outdated): “How To Write Industry-Dominating Meta Titles & Descriptions”

    After: “How to write title tags that Google will actually use”

    Yes, that last “before” was this very post. We all have a past.

    The better titles are specific, you can read them out loud without wincing, and they still make sense if the last word gets cut. They leave the details for the description and the first paragraph.

    Entity clarity, structured data, and social titles

    Search engines and answer engines have an easier time with a page that’s clear about who and what. Use the real organization name, the real service, and the real place in your title and opening text. Once, in natural language. Not a vague slogan repeated five times.

    Structured data won’t force a particular title link. And Google’s AI features documentation says you don’t need special schema to show up in AI Overviews or AI Mode.

    What structured data can do is confirm what’s already visible. Google’s introduction to structured data asks that markup match the text on the page. So the names and headlines in your Article or Organization markup should be the same ones a person can read.

    Social titles matter too. The Open Graph og:title and og:description tags, described in the Open Graph protocol, control how a link looks when it’s shared. Google also lists og:title as a source it may use for a title link.

    My advice? Keep your social title close to your search title. A dramatic social headline that doesn’t match the H1 creates the same kind of mismatch Google already rewrites.

    Writing titles for AI Overviews and other answer engines

    Google’s guidance on AI features lines up with the rest of Search: there’s no separate optimization track. To be eligible as a supporting link in AI Overviews or AI Mode, a page has to be indexed and eligible to show in Search with a snippet. A nosnippet rule, or max-snippet:0, can block that eligibility, as Google explains in its robots meta tag documentation.

    When it comes to the writing, I treat an AI answer the way I’d treat a busy reader.

    A model deciding whether to cite you will look at your title, your headings, and a short opening that states the point. If those lines are clear and agree with each other, you’re easier to quote accurately. If the title is stuffed, the H1 is cute, and the first paragraph takes six sentences to get going? You’re making the model (and the person) work way too hard.

    I won’t claim that ChatGPT search, Perplexity, Gemini, or Grok display your title tag in some documented, Google-like way. They don’t publish that. They cite pages they can crawl and parse.

    The practical work is the same anyway:

    • A precise title.
    • A matching H1.
    • An answer-first intro.
    • Visible text instead of text locked inside an image.
    • A description that could double as a summary.

    That’s also what Google’s note on succeeding in AI search keeps coming back to: useful, crawlable, people-first content, with preview controls that don’t needlessly restrict how your pages can be shown.

    How to test titles in Search Console

    After a round of title changes, I watch the affected URLs in the Search Console Performance report.

    Google’s guide to debugging Search traffic drops points out a handy pattern. If impressions stay roughly flat but clicks drop, your title and snippet may not be helping people understand the page.

    Compare a period after the recrawl with a similar period before, by page and by query. Read CTR as clicks divided by impressions, and check average position at the same time. If a title change lands alongside a big swing in rank, it’s not a clean test of your writing.

    Make one thoughtful change. Date it. Wait for the recrawl.

    And don’t show Google a different title than the one people see. Google’s website testing guidance is clear about cloaking, even during experiments.

    A practical checklist

    • Every indexable page has a unique, descriptive <title>. No “Home,” no site-wide slogan, no keyword list.
    • The important words come early enough to survive mobile truncation. The brand goes at the end, briefly, unless there’s a good reason to lead with it.
    • The title, the H1, and the first paragraph agree. The H1 is the clearest visual title on the page.
    • The title and the meta description do different jobs. They don’t contradict each other, and they’re unique to this URL.
    • The language and script match the page. Dates and claims in the title match what the page says now.
    • og:title lines up with the search title. Structured data, if you use it, matches the visible headline.
    • The page can be indexed and can show a snippet, if you want it eligible for AI Overviews or AI Mode.
    • After publishing, you watch the URL in Search Console (queries, impressions, clicks, and position) once Google has recrawled it.

    FAQ

    Why does Google show a different title than the one I wrote?

    Usually because the tag is vague, duplicated, outdated, stuffed, or out of step with your H1. Google lists these cases in its title-link guidance and may substitute other text. Fix the mismatch, request a recrawl, and wait.

    Should the title tag and the H1 be identical?

    They should mean the same thing. They don’t have to match word for word. Use the title tag for a tight, brand-aware headline in results, and the H1 for the on-page heading. Just don’t let them conflict on topic, date, or promise.

    How long should a title be?

    Google doesn’t set a character limit. The title link gets cut to fit the screen. I draft around 50 to 60 characters and put the distinctive words first. A slightly longer, accurate title beats a short, generic one.

    Do I need a special title format for AI search?

    No. Google says there are no extra requirements for AI Overviews or AI Mode beyond ordinary Search eligibility. Write a clear, unique title, line it up with the H1 and the intro, and keep the page snippet-eligible.

    The next step

    Open your highest-impression pages and compare three things: the title tag, the H1, and the title link Google is actually showing. Where they disagree, fix the writing, not the symptom. Then write the meta description so it finishes the thought the title started.

    If you’d like a second set of eyes on it, take a look at Seopya’s SEO and AI search services or get in touch.