Is AI going to kill SEO? Or just kill bad SEO?

Picture a Brattleboro marketing agency that spends a five-figure budget optimizing a client's website for traditional SEO. It stuffs keywords into every paragraph ("best plumber Brattleboro," "Brattleboro plumber services," "top-rated Brattleboro plumbing company"). It builds dozens of backlinks from low-quality directory sites and blog networks. It publishes 50 thin blog posts (300 words each, keyword-heavy, no real value). The site climbs to position three in Google's organic results for "plumber Brattleboro VT." The client gets some traffic and some leads. Then someone asks ChatGPT "Who are good plumbers in Brattleboro Vermont?" and the client is not mentioned. ChatGPT recommends three competitors instead, and one of them sits much lower in Google's organic results. The difference is not traditional SEO ranking. The difference is that the competitor has extractable, factual content (clear service descriptions, pricing, timelines) backed by third-party validation (detailed reviews, a quote in the local paper about frozen pipe prevention). It is also present on several platforms (Google, Yelp, Angi). AI tools trust and cite these signals. They ignore the keyword-stuffed content and the thin blog posts because they contain no useful facts AI can extract and synthesize.

This is the shift happening right now. AI search is not killing SEO. It is killing bad SEO (keyword stuffing, thin content, link manipulation, vague marketing language). Good SEO (clear facts, real authority, useful content that answers questions) is more important than ever. AI just makes it weigh more. The businesses that built their rankings on manipulation and fluff are watching their visibility collapse. The businesses that built their rankings on substance and value are thriving in the AI era because the same signals that helped them rank in Google (authoritative content, credible backlinks, user engagement) are the signals AI prioritizes when deciding who to recommend.

If you spent the last decade gaming Google with tricks and shortcuts, AI search is going to hurt. If you spent the last decade building genuinely useful content and real authority, AI search is going to help. The question is not whether AI is killing SEO. The question is which kind of SEO you have been doing.

What "bad SEO" looked like (and why it worked for a while)

Bad SEO was optimization focused on manipulating ranking algorithms rather than serving users. It worked because early search algorithms (Google's PageRank in the 2000s and early 2010s) were easier to game. If you knew the signals Google was looking for (keyword density, backlink count, exact-match domains, meta keywords), you could artificially boost those signals and rank well even if your content was mediocre.

Here are the tactics that defined bad SEO for the past 15 years:

Keyword stuffing: Writing content that repeats the target keyword unnaturally ("Looking for a Brattleboro plumber? Our Brattleboro plumbing company offers Brattleboro plumbing services for all your Brattleboro plumbing needs"). The goal was to hit a keyword density threshold (2% to 5% of total words) because early Google algorithms weighted keyword frequency heavily. The content was awkward and hard to read, but it ranked.

Thin content: Publishing dozens or hundreds of low-value pages (200 to 400 words each) targeting long-tail keywords. The content provided minimal value but ranked because Google counted pages and keywords, not quality. Example: a contractor publishing 50 separate pages for "kitchen remodel Brattleboro," "kitchen remodel Putney," "kitchen remodel Bellows Falls," each with the same 300-word template and just the town name swapped.

Link farms and PBNs (private blog networks): Building or buying backlinks from low-quality sites (directories, blog networks, link farms) to inflate link count. Early Google algorithms weighted backlink quantity heavily, so a site with 500 low-quality backlinks could outrank a site with 10 high-quality backlinks. The links provided no real value to users, but they gamed the algorithm.

Exact-match domains: Registering domains that exactly matched a target keyword (e.g., BrattleboroPlumber.com, BestVermontLawyer.com) because Google gave ranking weight to keywords in the domain name. The domain name did not reflect a real brand, but it ranked.

Hidden text and cloaking: Hiding keyword-stuffed text on the page (white text on white background, text sized at 1px, text hidden with CSS) so users did not see it but search engines indexed it. Or showing one version of the page to search engines (keyword-rich) and a different version to users (readable). This was outright manipulation and violated Google's guidelines, but it worked for years before Google got better at detecting it.

Meta keyword spamming: Stuffing the meta keywords tag (an HTML tag invisible to users but readable by search engines) with hundreds of keywords. Google stopped using meta keywords as a ranking signal in 2009, but many SEO practitioners kept doing it out of habit or ignorance.

These tactics worked for a while because Google's algorithm was simpler and more easily gamed. But over the past decade, Google has released a series of algorithm updates (Panda in 2011, Penguin in 2012, Hummingbird in 2013, RankBrain in 2015, BERT in 2019, plus ongoing core updates) designed to detect and penalize manipulative tactics. Those same updates reward high-quality, user-focused content. By 2020, most of the tactics above either stopped working or actively hurt rankings. Keyword stuffing triggers Google's spam filters. Link farms and PBNs get de-indexed. Thin content gets buried or filtered out. Exact-match domains lost their ranking boost.

But some businesses kept doing bad SEO because they did not know better, or because their SEO agency was stuck in 2010, or because they saw short-term gains and did not realize the long-term risk. Those businesses are now getting hit hard by AI search, because AI reads content even more critically than Google does. AI ignores fluff and extracts only facts. It also prioritizes authoritative sources. If your SEO strategy was built on manipulation, AI sees through it instantly.

Here is how that plays out. Say a Bellows Falls auto repair shop hired an SEO agency years ago, and the agency produced 60 thin blog posts (each 250 to 350 words, keyword-stuffed, topics like "why you need an oil change in Bellows Falls" and "best brake repair Bellows Falls VT"). The posts ranked in Google for a while and drove some traffic. Now the owner asks ChatGPT "Where should I get my brakes done in Bellows Falls?" and the shop is not mentioned. Instead it names two competitors with fewer blog posts and lower Google rankings. Looking at their sites, the owner finds detailed service pages with real information (what brake service includes, how long it takes, what it costs, when you need it). The shop's blog posts are vague and repetitive ("Brakes are important for safety. Get your brakes checked in Bellows Falls. Call us for brake service."). AI skips the keyword-stuffed posts and cites the factual service pages. The bad SEO that once worked is invisible to AI.

What "good SEO" looks like (and why it works even better in the AI era)

Good SEO is optimization focused on serving users and building real authority. It worked in traditional Google search because Google's algorithm evolved to prioritize user satisfaction (pages that answer questions, sites that load fast, content that users engage with). It works even better in AI search because AI is trained to extract useful information and cite authoritative sources. Good SEO and AI optimization are the same strategy.

Here are the pillars of good SEO:

Clear, factual, extractable content: Writing content that answers questions with specific facts (what, how, how much, how long). Example: instead of "We offer quality kitchen remodeling services in Brattleboro," write "We remodel kitchens in Brattleboro, Putney, and Bellows Falls. A typical kitchen remodel costs $18,000 to $35,000 depending on the scope (cabinets, countertops, flooring, appliances) and takes 4 to 8 weeks from design to completion. We handle design, permitting, demolition, installation, and final inspection." The second version is extractable. AI can pull facts (cost range, timeline, scope, service area) and cite them when someone asks a question.

Authoritative backlinks and mentions: Earning links and mentions from credible third-party sources (local news, industry publications, well-known directories, government sites, educational institutions). Example: getting quoted in a Brattleboro Reformer article about home energy efficiency, or getting listed on a Vermont state government contractor directory, or getting a backlink from a regional architecture blog. These signals tell Google (and AI) that other credible sources trust and reference you, which builds your authority.

Detailed, structured service pages: Building service pages that clearly describe what you do, how you do it, what it costs, and who you serve. Use structured headings (H2, H3), bullet lists, and clear sections. Example: a lawyer's estate planning page with sections for "What estate planning includes," "How the process works," "Typical timeline and cost," "Who needs estate planning," "Frequently asked questions." The structure makes the content easy to read (for humans) and easy to parse (for AI and Google).

Third-party validation: Building credibility through reviews, testimonials, case studies, certifications, awards, and partnerships. Example: collecting detailed Google and Yelp reviews, displaying industry certifications (licensed contractor, board-certified attorney, accredited business), listing professional associations (Vermont Bar Association, National Association of Home Builders), and showing client results (case studies, before/after photos, project timelines). These signals tell Google and AI that your claims are validated by independent sources, not just self-promotion.

User-focused design and performance: Building a fast, mobile-friendly, easy-to-navigate website that serves users first and search engines second. Example: a site that loads in under 2 seconds and works perfectly on mobile. It has a clear navigation menu with readable fonts and spacing, plus a simple contact form. Google prioritizes user experience signals (Core Web Vitals, mobile usability, engagement metrics). AI does not directly measure site speed, but a fast, clear site makes it easier for AI to crawl and extract content.

Topic depth and expertise: Writing comprehensive content that demonstrates deep knowledge of your subject. Example: instead of 20 thin blog posts (300 words each on surface-level topics), write 5 in-depth guides (1,500 to 2,500 words each covering a topic thoroughly with examples, step-by-step explanations, visuals, and FAQs). Google's algorithm (especially BERT and subsequent updates) prioritizes content that demonstrates expertise, authoritativeness, and trustworthiness (E-A-T). AI prioritizes the same signals.

These tactics have always been good SEO. The difference now is that AI makes them matter more. In traditional Google search, you could sometimes rank well with mediocre content if you had enough backlinks or keyword optimization. In AI search, you cannot. AI reads your content literally. It pulls out the facts and skips the fluff. If your content is vague or promotional, AI has nothing to cite. If your content is clear and factual, AI cites you.

Now the opposite case. Imagine a Putney CPA who built his website with good SEO principles. Clear service pages (tax preparation, bookkeeping, business consulting, estate planning). Each service page lists what is included, typical cost, timeline, who it is for, and how to get started. He publishes a handful of long-form guides on Vermont-specific tax topics (Vermont small business taxes, self-employment tax, LLC vs S-corp, home office deduction, estimated quarterly taxes), each detailed, factual, and full of examples. He is quoted in a local news article about Vermont tax filing deadlines. He collects detailed Google and Yelp reviews. When people ask ChatGPT or Perplexity about Vermont CPAs, tax preparation costs, or small business accounting, those tools have plenty to pull from: his service pages, his guides, the news article, and his reviews. The same signals (clear content, authoritative mentions, third-party validation) that help him rank in Google are the signals AI prioritizes.

Why AI reads content more literally than Google (and why fluff dies instantly)

Google's algorithm evaluates content using hundreds of signals (keywords, backlinks, user engagement, site structure, mobile usability, page speed, domain authority, etc.). It is a weighted, statistical model that tries to predict which pages will satisfy a user's query. This complexity created wiggle room for manipulation. You could rank with mediocre content if you optimized other signals well (backlinks, technical SEO, keyword placement).

AI search is different. AI reads your content literally. It parses the text, extracts facts, and synthesizes them into an answer. If your content contains no facts, AI has nothing to extract. If your content is vague or promotional ("We provide trusted, customer-focused plumbing solutions with a commitment to excellence"), AI cannot cite it because there is no specific information to pull.

Here is an example:

Bad SEO content (keyword-stuffed, vague):

"Looking for a Brattleboro electrician? Our Brattleboro electrical services company has been serving Brattleboro and surrounding areas for years. We offer quality electrical work, professional service, and competitive pricing. Trust our experienced Brattleboro electricians for all your electrical needs. Call today for a free estimate."

What Google sees: The page targets "Brattleboro electrician" (keyword appears 4 times in 50 words, decent keyword density). The content is thin but not spammy. If the site has decent backlinks and a complete Google Business Profile, it might rank in the top 10.

What AI sees: No extractable facts. What electrical services do they offer? What does "competitive pricing" mean? How many years have they been in business? What areas do they serve beyond "surrounding areas"? What credentials do they have? AI cannot answer any user query with this content because it contains no specifics. AI ignores the page.

Good SEO content (clear, factual, extractable):

"We provide residential and commercial electrical services in Brattleboro, Putney, Bellows Falls, and Wilmington. Our services include electrical panel upgrades, wiring and rewiring, lighting installation, outlet and switch repair, generator installation, and emergency electrical service. We have been licensed electricians in Vermont for 22 years. Service calls start at $95. Most panel upgrades cost $1,800 to $3,500 and take one to two days to complete. We are available for emergency service 24/7. Call (802) 555-1234 or fill out our contact form to schedule service."

What Google sees: The page targets "electrical services" and related terms. It includes location keywords (Brattleboro, Putney, Bellows Falls, Wilmington), service keywords (panel upgrades, wiring, lighting, generator), and signals expertise (22 years, licensed). The content is clear and useful. If the site has decent backlinks and user engagement, it will rank well.

What AI sees: Extractable facts everywhere. Services offered: panel upgrades, wiring, lighting, outlets, generators, emergency service. Service area: Brattleboro, Putney, Bellows Falls, Wilmington. Credentials: licensed in Vermont, 22 years in business. Pricing: service calls start at $95, panel upgrades cost $1,800 to $3,500. Timeline: one to two days for panel upgrades. Availability: 24/7 emergency service. Contact: phone number and contact form. AI can answer multiple user queries with this content ("What does an electrical panel upgrade cost in Brattleboro?" → "$1,800 to $3,500 according to (business name)" / "Who offers emergency electrical service in Putney?" → "(Business name) offers 24/7 emergency service"). AI cites the page frequently because it is packed with useful, extractable facts.

The shift is brutal for businesses that relied on vague, keyword-heavy content. That content might still rank in Google (for now), but it is invisible to AI. AI needs facts to synthesize and cite. If your content does not provide facts, AI moves on to competitors who do.

For example, say a Wilmington HVAC contractor has a homepage written years ago. The content is vague and keyword-heavy ("Wilmington's trusted HVAC experts. Quality heating and cooling solutions. Professional service you can count on. Call Wilmington's #1 HVAC company today."). The page ranks fourth in Google organic search for "HVAC Wilmington VT" because the site has decent backlinks and an old domain (domain age was a minor ranking signal). Someone asks Perplexity "What does furnace installation cost in Wilmington Vermont?" and Perplexity does not mention him. It cites competitors who state clear pricing and timelines on their sites ("Furnace installation costs $4,200 to $6,800 depending on the unit" and "Most furnace installs are completed in one day"). His vague homepage provides no pricing, no timeline, no specifics, so there is nothing to extract. Rewriting the homepage with clear service descriptions, pricing ranges, timelines, and credentials is what gives AI something to cite. The factual rewrite makes his content AI-extractable.

Why thin content and keyword stuffing are dying faster in AI search than in Google

Google has been fighting thin content and keyword stuffing for over a decade (Panda and Penguin updates targeted these tactics starting in 2011 and 2012). But Google's enforcement has been uneven. Some keyword-stuffed pages still rank if they have strong backlinks or domain authority. Some thin content still appears in search results if it targets low-competition keywords.

AI search has zero tolerance for thin content and keyword stuffing because AI evaluates content differently. AI does not care about keyword density or domain age. It reads the text, extracts facts, and evaluates whether the content answers the query. If the content is thin (no depth, no facts, no value), AI has nothing to extract. If the content is keyword-stuffed (repetitive, awkward, focused on ranking instead of answering), AI sees it as low-quality and deprioritizes it.

Here is why thin content fails in AI search:

Thin content lacks extractable facts. A 300-word blog post that says "Estate planning is important. You should make a will. Contact us to discuss your estate planning needs" provides zero facts AI can extract. What does estate planning include? How much does it cost? How long does it take? Who needs it? What is the process? Thin content does not answer any of these questions, so AI cannot cite it.

Thin content lacks depth and context. AI synthesizes information from multiple sources and builds a comprehensive answer. If your content is shallow, AI pulls from deeper sources (competitors with longer, more detailed content). A 2,000-word guide that explains estate planning in detail (what it includes, how the process works, typical costs and timelines, who needs it, FAQs) is far more likely to be cited than a 300-word thin page that just says "we do estate planning."

Thin content signals low expertise. Google's E-A-T (expertise, authoritativeness, trustworthiness) framework prioritizes content that demonstrates deep knowledge. AI applies the same logic. A business that publishes 50 thin blog posts signals they are optimizing for keywords, not educating users. A business that publishes 10 comprehensive guides signals they have real expertise. AI trusts and cites the latter.

Here is why keyword stuffing fails in AI search:

Keyword stuffing is unnatural and hard to extract. AI parses sentences and extracts meaning. If a sentence is stuffed with keywords ("Our Brattleboro plumbing company offers Brattleboro plumbing services including Brattleboro drain cleaning and Brattleboro pipe repair"), the sentence is awkward and the meaning is buried. AI struggles to extract clean facts from keyword-stuffed text because the repetition creates noise.

Keyword stuffing signals low quality. AI is trained on high-quality text (news articles, Wikipedia, authoritative publications). Keyword-stuffed content does not match the quality patterns AI recognizes. AI deprioritizes it as low-quality or spammy.

Keyword stuffing does not answer questions. Users ask AI questions in natural language ("What does drain cleaning cost in Brattleboro?"). AI looks for content that answers the question clearly and directly. Keyword-stuffed content optimizes for ranking, not for answering questions. AI ignores it.

Consider a Londonderry real estate agent with 40 blog posts published by an SEO agency. Each post is 300 to 400 words, keyword-stuffed, and thin. Examples: "Londonderry VT real estate market update," "Best time to sell your Londonderry home," "Londonderry home values rising." Each post repeats the target keyword over and over and provides minimal useful information ("The Londonderry real estate market is strong. Now is a great time to sell your Londonderry home. Contact our Londonderry real estate team for a free market analysis."). The posts rank in Google for low-competition keywords and drive some traffic. But when someone asks ChatGPT "What are home prices in Londonderry Vermont?" or "When is the best time to sell a house in Vermont?" it has no reason to cite any of them. It will reach for listing-site data or a news article about Vermont housing trends. Or it cites a competitor's detailed market report with charts, specific price data, neighborhood breakdowns, and analysis. Thin, keyword-stuffed posts have no facts to extract, so AI ignores them. Replacing the 40 thin posts with a few in-depth guides (market trends with data, the home selling process, the home buying process, financing options, local neighborhood guides, seasonal market analysis) gives AI something worth citing. The shift from thin to deep content is what makes content AI-worthy.

Why link farms and low-quality backlinks are worthless in AI search

Backlinks (links from other websites to your site) are one of Google's core ranking signals. A site with many high-quality backlinks (from authoritative, relevant sites) ranks higher than a site with few backlinks. This created an industry of link-building services, many of which used manipulative tactics (link farms, private blog networks, paid links, link exchanges) to artificially inflate backlink counts.

Google has been fighting low-quality link schemes for years (the Penguin update in 2012 specifically targeted manipulative links). But enforcement is imperfect. Some sites with low-quality backlinks still rank because Google's algorithm cannot always detect manipulation or because the other ranking signals (content quality, user engagement) are strong enough to compensate.

AI search does not care about backlink count. AI cares about content quality and source authority. Here is how AI evaluates sources:

AI prioritizes authoritative domains. When AI searches the web to answer a query, it weights sources based on perceived authority. Trusted sources include news sites (New York Times, local newspapers) along with .gov and .edu domains. Well-known industry publications and established platforms (Wikipedia, Healthgrades, Yelp) count as well. Random blogs, low-quality directories, and link farm sites are ignored or deprioritized.

AI evaluates content quality, not link count. If your site has 500 backlinks from low-quality sources (directories, blog networks, link farms) but your content is thin or vague, AI does not care about the backlinks. AI reads your content, finds no useful facts, and moves on. If a competitor has 10 backlinks from high-quality sources (a news article, a university blog, a government site) and clear, factual content, AI cites the competitor.

AI synthesizes from multiple independent sources. When AI builds an answer, it pulls from multiple sources and cites them. If all your backlinks are from a private blog network (a cluster of sites you control or pay for), AI sees those as non-independent sources and deprioritizes them. If your backlinks are from independent, credible sources (a local newspaper, an industry publication, a customer review platform), AI sees those as validation and cites them.

The practical implication: if you spent money on link-building services that delivered hundreds of low-quality backlinks, those links are worthless in AI search. They might help your Google rankings (marginally), but they do not help your AI visibility. If you earned a few high-quality backlinks (a news mention, a government directory listing, a link from a university or industry association), those links carry significant weight in AI search because they signal authority and credibility.

Say a Brattleboro law firm pays a link-building service, which delivers 200 backlinks from low-quality sites (generic legal directories, blog networks, unrelated blogs). The links lift the site's domain authority score (a metric used by SEO tools) and may contribute to some ranking gains in Google. But when someone asks ChatGPT about estate planning and family law in Brattleboro, it cites two competitors instead. Their backlink profiles are much smaller, but the links come from authoritative, relevant sources: a news article, the state bar association, a local nonprofit they partner with, a regional legal blog. AI weighs that kind of credibility heavily, and 200 low-quality links add little. The better strategy is to stop paying for link-building and earn credible mentions instead. Pitch local news and write guest posts for legal blogs. Join industry associations and partner with nonprofits. A handful of high-quality mentions does more than hundreds of weak links. The shift from quantity to quality makes the difference.

Why authoritative mentions (news, industry publications, directories) matter more in AI search

In traditional SEO, authoritative backlinks helped your Google rankings. In AI search, authoritative mentions do more than help rankings: they get cited directly. When AI synthesizes an answer, it often pulls from and cites the authoritative source itself (the news article, the industry publication, the government directory). If you are mentioned in that source, you get cited by association.

Here is how authoritative mentions work in AI search:

News mentions: If you are quoted in a local news article (Brattleboro Reformer, the Commons, Eagle Times, VTDigger), that article gets indexed by AI. When someone asks a question related to the article's topic, AI may cite the article and mention you by name. Example: a Bellows Falls contractor is quoted in an Eagle Times article about rising construction costs in Vermont. Two months later, someone asks ChatGPT "Why are construction costs going up in Vermont?" ChatGPT cites the Eagle Times article and mentions the contractor as a source.

Industry publications: If you write a guest article for a trade publication or you are quoted in an industry blog, AI indexes that content and cites it when relevant. Example: a Vermont CPA writes a guest post for a regional accounting newsletter about Vermont tax law changes. Perplexity cites the article when someone asks about Vermont business taxes.

Government and .edu sites: If your business is listed on a government site (a state contractor directory, a licensing board or a town site) or an educational site (a university extension service, a community college business directory), AI weights that mention heavily because .gov and .edu domains are inherently trusted.

Industry associations and certifications: If you are listed on a professional association website (Vermont Bar Association, National Association of Home Builders, Better Business Bureau accreditation), AI treats that as a credibility signal.

The strategy: focus on earning mentions from authoritative, independent sources. Pitch local news. Write guest posts for industry blogs. Get listed on government directories. Join industry associations and make sure you are listed on their member directory. These mentions build your authority in Google search and in AI search simultaneously.

A Westminster roofer was quoted in a Brattleboro Reformer article about preparing Vermont homes for winter (the article covered gutter cleaning, roof inspections, ice dam prevention). The roofer provided specific tips and cost estimates. The article was published online. Over the next four months, ChatGPT cited the article 6 times when users asked about roof maintenance and gutter cleaning costs. Ice dam prevention questions in Vermont brought it up as well. Each time, ChatGPT mentioned the roofer by name and included a link to his website. The one news mention drove an estimated 4 to 5 new project inquiries (tracked via referral traffic from ChatGPT and customer intake forms). The authoritative mention (Reformer article) carried more weight in AI search than the roofer's entire website.

How to audit your SEO strategy for AI readiness (the diagnostic)

If you have been doing SEO for years and you are not sure whether your strategy is "good SEO" or "bad SEO" in the AI era, run this diagnostic:

Step one: Check for keyword stuffing

Open your homepage and your top three service pages. Read them aloud. Do they sound natural, or do they sound like you are repeating keywords unnaturally? If you say the same phrase (e.g., "Brattleboro plumber") more than once or twice in a single paragraph, you are probably keyword stuffing. Rewrite for clarity and natural language.

Step two: Check for thin content

Look at your blog archive or your service pages. How many pages have fewer than 500 words? How many of those pages provide real depth and value, versus surface-level fluff? If you have 30 blog posts that are each 300 words and provide no new information (just keyword variations), that is thin content. Consider consolidating or deleting them and replacing them with 5 to 10 in-depth guides (1,500+ words each).

Step three: Check for extractable facts

Open your homepage and your main service page. Can you identify specific facts an AI could extract? Look for numbers (pricing, timelines, years in business), lists (services offered, towns served), credentials (licenses, certifications), and process descriptions (how the work gets done, step-by-step). If your content is vague ("quality service," "trusted expertise," "customer-focused solutions"), it is not extractable. Rewrite with specifics.

Step four: Check your backlink profile

Use a free backlink checker (Moz, Ahrefs, SEMrush all offer limited free checks, or use Google Search Console). Look at your top 20 to 30 backlinks. How many are from authoritative, relevant sources (news sites, industry publications, government sites, well-known directories)? How many are from low-quality sources (random blogs, link farms, unrelated sites)? If most of your backlinks are low-quality, they are not helping you in AI search. Shift focus to earning high-quality mentions.

Step five: Test your AI visibility

Ask ChatGPT, Perplexity, and Claude three to five queries related to your business and location. Are you cited? If yes, which sources does the AI pull from (your website, your reviews, a news article, a directory)? If no, which competitors are cited and which sources does AI pull from for them? This tells you where your gaps are.

Step six: Compare your content to competitors

Visit the websites of the competitors AI cited. Read their service pages and blog posts. Is their content more detailed than yours? More factual? Better structured? Do they have pricing, timelines, process explanations, FAQs? If their content is clearly more useful and extractable than yours, that is your gap.

A Grafton accountant ran this diagnostic in early 2025. Step one: his homepage was keyword-stuffed (the phrase "Grafton CPA" appeared 8 times in 250 words). Step two: he had 22 blog posts, 18 of which were under 400 words and provided minimal value. Step three: his service pages were vague (no pricing, no timelines, no process descriptions). Step four: he had 140 backlinks, 110 of which were from low-quality directories. Step five: he tested ChatGPT and was cited zero times out of five queries. Competitors with fewer backlinks and lower Google rankings were cited because they had clear pricing and detailed service pages. They also had credible mentions (one had a VTDigger quote, another had detailed Yelp reviews). Step six: the competitors' content was more factual and more structured than his. The diagnostic identified the gaps: keyword-stuffed content, thin blog posts, vague service pages, low-quality backlinks. The accountant spent two months fixing the gaps (rewrote homepage and service pages with clear facts; replaced thin blog posts with 4 in-depth tax guides; got quoted by VTDigger; worked on reviews). He tested ChatGPT again. Now he was cited in three out of five queries. The diagnostic and the fixes transformed his AI visibility.

How to shift from bad SEO to good SEO in 90 days (the action plan)

If your diagnostic revealed that your SEO strategy leans toward "bad SEO" (keyword stuffing, thin content, low-quality backlinks), here is the 90-day plan to fix it:

Month one: Rewrite your core pages for clarity and extractability

Week one: Rewrite your homepage. Remove keyword stuffing. Lead with a clear statement of what you do, who you serve, and where you are located. Add a bulleted service list and a town list. Then add a brief credentials section (years in business, licenses, certifications). Add a clear call-to-action (contact form, phone number, or "get started" link).

Week two: Rewrite your top three service pages. For each service, say what it is and what it includes. Say who it is for and how long it takes. Give what it costs (at least a range) and how the process works. Finish with the credentials or experience you bring. Use headings (H2, H3), bullet lists, and short paragraphs. Cut vague marketing language.

Week three: Add or update your FAQ page. Write 10 to 15 questions your customers actually ask (you can pull these from customer emails, phone calls, or Google's "People also ask" box). Answer each question with specific facts (not fluff).

Week four: Review your blog. Delete or consolidate thin posts (anything under 500 words that provides no unique value). Identify 3 to 5 topics you can write comprehensive guides on (1,500+ words each, based on customer questions or industry expertise). Start drafting the first guide.

Month two: Earn one authoritative mention and improve your third-party presence

Week one: Claim and optimize your profiles on Yelp and one industry-specific platform (Avvo, Houzz, Healthgrades, Angi, etc.). Complete every section, upload photos, respond to existing reviews.

Week two: Ask every customer for a review on Google, Yelp, or the industry platform. Give them a gentle prompt to include details (service type, timeline, result). Goal: collect 5 to 10 detailed reviews this month.

Week three: Pitch a local news story or write a guest post. Identify a newsworthy angle (a project, a milestone, a trend, an expert perspective). Write a pitch email and send it to a local reporter or a regional publication. (See the citations and mentions post for pitch templates.)

Week four: Answer three Q&A questions on a relevant platform (Avvo, Quora, Reddit, industry forum). Write detailed, factual answers (300+ words each) that demonstrate expertise and mention your location naturally.

Month three: Publish in-depth content and test your AI visibility

Week one: Publish the first comprehensive guide (1,500+ words) on your blog or as a standalone resource page. Promote it on social media and in your email newsletter (if applicable).

Week two: Publish the second comprehensive guide. Cross-link the two guides where relevant topics overlap.

Week three: Check whether the news pitch or guest post from month two was published. If yes, link to it from your website (add a "Press" or "In the news" section to your About page or homepage). If no, follow up with the reporter or publication, or pitch a second outlet.

Week four: Test your AI visibility. Ask ChatGPT, Perplexity, and Claude your core queries. Are you cited now? Which sources do they pull from? Compare to your baseline from the diagnostic. Note improvement and identify remaining gaps.

A Townshend general contractor followed this plan starting in February. Month one: rewrote homepage, services pages, and FAQ with clear facts (service list, town list, pricing ranges, timelines, credentials), deleted 18 thin blog posts. Month two: claimed Yelp and Houzz, collected 9 reviews, pitched the Commons (they ran a story about a historic barn restoration he completed), answered 3 construction questions on Houzz. Month three: published two in-depth guides (one on remodeling costs and timelines in Vermont, one on building in Vermont's climate), linked to the Commons article from his homepage, tested AI visibility. Result: ChatGPT cited him in three out of five queries (up from zero out of five before the plan). Perplexity cited him in four out of five. The 90-day shift from bad SEO to good SEO transformed his AI visibility and his credibility.

Why "good SEO" and "AI optimization" are the same strategy

The clearest proof that AI is not killing SEO (just killing bad SEO) is this: the tactics that work best for AI visibility are the same tactics that work best for modern Google SEO. There is no conflict. Good SEO is AI optimization.

Here is the overlap:

Clear, factual content: Google's BERT and RankBrain updates prioritized content that answers questions clearly and naturally. AI search does the same. Clear, factual content that answers questions ranks well in Google and gets cited by AI.

Authoritative backlinks and mentions: Google's PageRank algorithm (and all subsequent updates) prioritizes links from authoritative, relevant sites. AI search prioritizes mentions and citations from authoritative sources. If you earn high-quality backlinks (news mentions, industry publications, credible directories), you rank well in Google and you get cited by AI.

User-focused structure and performance: Google's Core Web Vitals and mobile-first indexing prioritize fast, mobile-friendly, easy-to-navigate sites. AI does not directly measure site speed, but a well-structured, fast site makes it easier for AI to crawl and extract content. If you build a user-focused site, you rank well in Google and you make your content more accessible to AI.

Third-party validation: Google's E-A-T framework prioritizes content and sites that demonstrate expertise, authoritativeness, and trustworthiness. AI search prioritizes third-party validation (reviews, mentions, Q&A answers, news quotes). If you build third-party credibility, you rank well in Google and you get cited by AI.

Topic depth and comprehensiveness: Google's Hummingbird and subsequent updates prioritize comprehensive content that covers a topic in depth. AI search prioritizes deep content because it provides more extractable facts and demonstrates expertise. If you write in-depth guides, you rank well in Google and you get cited by AI.

The strategy is the same. The businesses that built their SEO on substance (clear content, real authority, user focus) are winning in both Google search and AI search. The businesses that built their SEO on manipulation (keyword stuffing, thin content, link schemes) are losing in both.

Imagine a Newfane lawyer who built her website with good SEO principles from the start (clear service descriptions, in-depth guides on Vermont estate planning and family law, authoritative backlinks from her professional association and a local news quote, detailed client reviews on Google and Avvo). She ranks well in Google for most of her target keywords. When AI search arrives, she does not need to change her strategy, because AI tools have what they need to cite her: her service pages, her guides, the news article, and her Avvo profile. Her strategy was already aligned with what AI prioritizes (clear facts, authoritative mentions, third-party validation, depth). Good SEO becomes good AI optimization with little additional effort.

What Google's algorithm updates over the past decade tell us about the future

Google has released a series of major algorithm updates over the past 15 years, each targeting a specific type of manipulation or low-quality content. These updates provide a roadmap for where search (and AI search) is heading:

Panda (2011): Targeted thin content, duplicate content, and low-quality content farms. Message: content must provide unique value, not just keyword variations.

Penguin (2012): Targeted manipulative link schemes (link farms, paid links, low-quality directories). Message: backlinks must be earned from credible sources, not bought or manufactured.

Hummingbird (2013): Shifted Google from keyword matching to semantic search (understanding the meaning and intent behind queries, not just matching keywords). Message: write for users and intent, not for keyword density.

Mobile-first indexing (2015 onward): Prioritized mobile-friendly sites and fast-loading pages. Message: user experience (especially on mobile) is a ranking signal.

RankBrain (2015): Introduced machine learning to Google's algorithm, allowing it to better understand query intent and user satisfaction. Message: Google is learning which results satisfy users, and it adjusts rankings based on engagement.

BERT (2019): Improved Google's understanding of natural language and context (especially for long, conversational queries). Message: write naturally, answer questions clearly, and provide context.

Core Web Vitals (2021): Made page speed, interactivity, and visual stability explicit ranking factors. Message: site performance and user experience matter.

Helpful Content Update (2022 and ongoing): Targeted content written for search engines rather than users (content that exists only to rank, not to help). Message: write for humans first, search engines second.

The pattern is clear: Google has spent 15 years moving away from keyword-based ranking and toward intent-based, user-focused ranking. AI search is the logical next step in this evolution. AI reads content the way a human would, extracts meaning and facts, and prioritizes sources that provide real value. The businesses that adapted to Google's algorithm updates (by shifting from keyword stuffing to clear writing, from link schemes to earning credible mentions, from thin content to comprehensive guides) are the businesses that are thriving in AI search. The businesses that ignored the updates and kept doing bad SEO are the businesses that are invisible to AI.

Consider a Brattleboro retail store owner who hired an SEO agency in the early 2010s. The agency stuffed keywords and built low-quality backlinks. It published thin content too. The site ranked well until the Panda update in 2011 (rankings dropped). The agency added more backlinks. Rankings recovered slightly until the Penguin update in 2012 (rankings dropped again). The agency kept doing the same things (more keywords, more thin content, more junk links). The site sank. If the owner then switches to an approach focused on good SEO (rewriting content for clarity and value, earning credible backlinks, building third-party presence, publishing comprehensive guides), rankings can recover over time, and that same work is what makes the store citable by ChatGPT and Perplexity today. The shift from bad SEO to good SEO protects visibility in both Google search and AI search.

Why businesses that ignored Google's updates are getting hit hardest by AI search

If you ignored Google's algorithm updates over the past decade (you kept keyword stuffing, kept publishing thin content, kept buying low-quality backlinks), your site probably lost Google rankings gradually over time. You might have seen traffic decline year over year, or you might have plateaued while competitors passed you. AI search accelerates that decline because AI has zero tolerance for the tactics Google has been penalizing for years.

Here is why businesses that ignored Google's updates are struggling in AI search:

They still have keyword-stuffed content. Google deprioritizes keyword stuffing (it triggers spam filters and reduces rankings). AI ignores keyword-stuffed content entirely because it is unnatural and hard to extract facts from. If your content is still stuffed with keywords, you are invisible to AI.

They still have thin content. Google's Helpful Content Update and Panda update penalize thin, low-value content. AI ignores thin content because it lacks depth and extractable facts. If you have 50 thin blog posts and no comprehensive guides, AI has nothing to cite.

They still have low-quality backlinks. Google's Penguin update penalizes manipulative link schemes. AI ignores low-quality backlinks because it prioritizes authoritative sources. If your backlink profile is full of link farm sites and blog networks, those links do nothing for your AI visibility.

They have not built third-party presence. Google's E-A-T framework prioritizes third-party validation (reviews, mentions, credentials). AI prioritizes the same signals. If you focused only on your website and ignored third-party platforms (Yelp, industry sites, news mentions), you have no third-party signals for AI to cite.

The businesses that adapted to Google's updates (by rewriting content for clarity, earning credible backlinks, building third-party presence, publishing in-depth guides) did the work AI search requires before AI search existed. They are ready. The businesses that ignored Google's updates are now paying double: they rank poorly in Google and they are invisible to AI.

Or picture a Putney home services business that ignored Google's updates for most of a decade. It kept keyword stuffing and kept publishing thin blog posts (dozens of them, averaging under 300 words). It also kept buying backlinks from low-quality directories. Its Google rankings slid steadily. Then a new marketing person rebuilds the strategy. All content gets rewritten for clarity and depth, and most of the thin posts give way to a few comprehensive guides. Backlink buying stops. The focus shifts to earning reviews and news mentions. Rankings can recover over the following year, and the business becomes citable by ChatGPT and Perplexity because its content is now clear and factual, with third-party validation (a news mention, Yelp reviews, a Houzz profile) behind it. The businesses that did the work to adapt to Google's updates were ready for AI. The ones that kept doing bad SEO are scrambling now.

How to explain this shift to a boss or a client who is panicking about AI

If you are a marketer or an in-house SEO professional and your boss or client is panicking about AI search ("Is AI going to kill our SEO? Should we stop investing in SEO? What do we do?"), here is how to explain the shift:

AI is not killing SEO. AI is accelerating the shift Google started a decade ago. Google has been penalizing bad SEO (keyword stuffing, thin content, link schemes) since 2011. AI search has zero tolerance for those tactics. If we have been doing good SEO (clear content, authoritative backlinks, third-party presence, in-depth guides), we are already aligned with what AI prioritizes. If we have been doing bad SEO, we need to fix it, but the fixes are the same fixes Google has been asking for since 2011.

The businesses that are invisible to AI are the ones that never adapted to Google's updates. If we adapted to Panda (no thin content), Penguin (no link schemes), BERT (clear, natural writing), and Helpful Content (user-focused, not keyword-focused), we are ready for AI. If we did not adapt, we have work to do, but the work is not new: it is catching up on what Google has been signaling for years.

The tactics that work for AI visibility are the same tactics that work for Google rankings. Clear, factual content. Authoritative backlinks and mentions. Third-party validation (reviews, news, Q&A). In-depth, comprehensive guides. User-focused site design. We are not adding a new strategy on top of SEO. We are refining the SEO strategy we should have been doing all along.

AI actually rewards businesses that have been doing good SEO. If we have been earning credible backlinks, collecting detailed reviews, publishing in-depth content, and building third-party presence, AI cites us more often than competitors who have been doing bad SEO. AI search is an opportunity for businesses that did the work.

The action plan is straightforward. Audit our content for keyword stuffing and thin pages (rewrite or delete). Audit our backlinks for low-quality sources (disavow or ignore them, focus on earning high-quality mentions). Audit our third-party presence (claim profiles, collect reviews, answer Q&A questions, pitch news). Publish in-depth guides on topics our customers care about. Test our AI visibility monthly and adjust. This is good SEO. It works for Google and it works for AI.

This framing reassures the boss or client (AI is not a doomsday scenario) and focuses them on actionable steps (audit and fix bad SEO, double down on good SEO).

Why businesses that invested in good SEO are seeing an AI visibility boost

If you spent the past five years investing in good SEO (clear content, credible backlinks, third-party presence, in-depth guides, user-focused design), you are probably seeing better AI visibility than you expected. Here is why:

Your content is already extractable. You wrote for users, not for keywords, so your content is clear and factual. AI can read it, extract facts, and cite it. Competitors who wrote for keywords have vague, repetitive content AI ignores.

Your backlinks are from credible sources. You earned links from news sites, industry publications, and well-known directories. AI trusts and cites those sources. Competitors who bought links from link farms have backlinks AI ignores.

Your third-party presence is strong. You collected reviews on multiple platforms (Google, Yelp, industry sites), you answered Q&A questions, you got mentioned in local news. AI synthesizes from those third-party sources and cites you. Competitors who focused only on their website have no third-party signals for AI to pull from.

Your content demonstrates depth and expertise. You published comprehensive guides that answer questions thoroughly. AI prioritizes deep content because it provides more facts and demonstrates authority. Competitors who published thin blog posts have nothing AI can cite.

The investment you made in good SEO is paying off in AI search with zero additional effort. The businesses that cut corners are now scrambling to catch up.

Think of a Bellows Falls financial planner who invests in good SEO for years. He publishes comprehensive guides on financial planning topics (retirement planning, college savings, estate planning, tax-efficient investing, insurance, Social Security). He is quoted in a local news article about Vermont retirement trends. He collects detailed reviews on Google and Yelp. He answers financial planning questions on Quora and Reddit. His website is fast and clearly structured. He ranks well in Google for most of his target keywords. When people ask ChatGPT, Perplexity or Claude about those topics, the tools have plenty to draw on: his guides, the news article, his reviews, and his Q&A answers. He does not need to change anything when AI search emerges, because his good SEO strategy is already aligned with what AI prioritizes. Competitors who did bad SEO (thin content, keyword stuffing, low-quality backlinks) are the ones struggling to catch up.

The bottom line: AI rewards substance, not shortcuts

AI search is not killing SEO. It is killing the shortcuts and manipulation that never should have worked in the first place. Keyword stuffing was always a bad user experience (awkward, hard to read, focused on gaming algorithms instead of helping people). Thin content was always low-value (surface-level fluff that answered no real questions). Link farms and PBNs were always manipulation (artificial signals that did not reflect real authority). These tactics worked for a while because early search algorithms were easier to game. But Google spent 15 years closing the loopholes, and AI search has zero tolerance for them.

Good SEO (clear facts, real authority, useful content) has always been the right strategy. It worked in 2010. It works in Google search today and in AI search. The businesses that built their strategy on substance are thriving. The businesses that built their strategy on shortcuts are watching their visibility collapse.

If you have been doing bad SEO, the fix is not complicated. Stop stuffing keywords and start writing clearly. Stop publishing thin content and start publishing in-depth guides. Stop buying backlinks and start earning credible mentions. Stop focusing only on your website and start building third-party presence (reviews, Q&A, news mentions). The tactics are the same tactics Google has been rewarding since 2011. They work for Google, and they work for AI.

If you have been doing good SEO, keep doing it. Your content is already AI-ready. Your backlinks are already credible. Your third-party presence is already strong. AI search is an opportunity, not a threat. The businesses that did the work are winning.

AI is not killing SEO. AI is just making the rules count harder. Build for substance instead of shortcuts and you will be visible in every search environment that matters.

Want to see whether your SEO strategy is AI-ready? Run the diagnostic above. Check your pages for keyword stuffing and thin content, and compare your backlink quality to competitors. Ask ChatGPT and Perplexity the questions your customers ask. Then build a 90-day plan for shifting from bad SEO to good SEO. If you need a clear, factual website as the foundation, Brattleboro Local offers a simple one: six pages written for you from what you tell us (including a page of answers to common questions), on your own web address, for $499 a year. See the details and price.