Perplexity, ChatGPT, Claude: which one are your customers actually using?

Imagine a Brattleboro dentist who optimizes her website for ChatGPT. She runs queries, confirms ChatGPT cites her, and feels confident. Then a prospective patient asks Perplexity "Who are good cosmetic dentists in Brattleboro Vermont?" and gets a list of recommendations. She is not on it, and several of the practices listed are ones ChatGPT never mentioned. The patient books a consultation with one of Perplexity's picks. The dentist loses the lead because she optimized for one AI tool and ignored the others.

This is the trap most businesses are falling into. They test ChatGPT to see whether they are cited and assume that is enough. But your customers are not all using ChatGPT. Some are using Perplexity (especially for research-heavy or shopping queries). Some are using Claude (especially professionals and knowledge workers). Each tool pulls from different sources and weights signals differently. Each also serves a different user base. If you show up in one but not the others, you are invisible to a significant chunk of your potential customers.

The three major AI search tools (ChatGPT, Perplexity, and Claude) have distinct user bases, use cases, and ranking behaviors. ChatGPT dominates by volume (it has 200+ million active users and is the default AI tool for most people). Perplexity is growing fast for research, shopping, and local queries (it is designed specifically for search and discovery, and it surfaces real-time results with citations). Claude is favored by professionals, researchers, and people who want deeper analysis or more nuanced answers (it is known for careful reasoning and less hype). Each tool has strengths and weaknesses, and each one requires slightly different optimization strategies.

This post breaks down the three tools and who uses them and why. It shows how they differ in what they cite and recommend, and what you need to do to show up in all three. Because if your business is visible in only one of them, you are leaving money on the table.

ChatGPT: the volume leader, the general-purpose default

ChatGPT (from OpenAI) is the most widely used AI tool by a huge margin. As of late 2024, ChatGPT has over 200 million weekly active users. For comparison, Perplexity has around 10 million monthly users, and Claude (from Anthropic) does not publish user numbers but is estimated to be in the low millions. ChatGPT is the default AI for most people because it was first to market (November 2022) and it is free (with a paid tier for advanced features). It is also general-purpose (you can ask it anything, from writing help to coding to research to local business recommendations).

Who uses ChatGPT?

Everybody. ChatGPT has the broadest user base of any AI tool. Students use it for homework help. Office workers use it to draft emails and summarize documents. Small business owners use it to write website copy and marketing content. Consumers use it to plan trips, research products, and find local services. Developers use it to write code. Writers use it to brainstorm and edit. ChatGPT is the Swiss Army knife of AI tools, which means it attracts users across every demographic and use case.

What do people use ChatGPT for?

ChatGPT is most commonly used for:

Writing and editing: Drafting emails, blog posts, social media captions, cover letters, resumes, website copy. (This is probably the single most common use case.)

Learning and explanation: Asking "explain X to me" or "how does Y work" or "what is the difference between A and B." ChatGPT is good at breaking down complex topics into plain language.

Brainstorming and ideation: Generating ideas, outlines, lists, project names, marketing angles.

Coding and technical help: Writing code snippets, debugging, explaining technical concepts.

Local business and service recommendations: "Who are good plumbers in Brattleboro Vermont?" or "Where should I eat in Bellows Falls?" This use case is growing fast as more people realize ChatGPT can answer local questions (it was not always good at this, but it has improved significantly in 2024).

General knowledge and research: Quick answers to factual questions, though ChatGPT is less reliable for this than Perplexity (because ChatGPT does not always cite sources and its training data has a cutoff date, so it can be outdated on recent events).

How does ChatGPT decide who to recommend?

When you ask ChatGPT for local business recommendations ("Who are good contractors in Putney Vermont?"), it synthesizes an answer by pulling from its training data (which includes web pages, reviews, directories, news articles indexed before its cutoff date) and, if you are using the paid version (ChatGPT Plus or Enterprise), real-time web search (it can search the live web and pull current results). The free version relies on training data, which means it can be outdated. The paid version is more current.

ChatGPT weights several signals when deciding who to recommend:

Review volume and quality: Businesses with many detailed reviews on Google, Yelp, and other platforms get cited more often. ChatGPT extracts facts from reviews (what service was provided, how it went, what the outcome was). It uses those facts to build its recommendations.

Website content clarity and depth: Businesses with clear, factual website content (services, pricing, timelines, credentials, process descriptions) get cited more often than businesses with vague marketing language. ChatGPT extracts facts from websites (it reads your homepage, your service pages, your About page). Then it synthesizes them into its answer.

Third-party mentions: Businesses mentioned in news articles, blogs, directories, and Q&A platforms (Quora, Reddit, Avvo, Houzz) get cited more often. ChatGPT treats third-party mentions as validation signals (if multiple independent sources mention you, you are more credible).

Authority and credentials: Businesses that display credentials (licenses, certifications, years in business, professional associations) are weighted as more authoritative. ChatGPT looks for verifiable expertise signals (see the E-E-A-T post for more on this).

Recency and activity: Businesses that are active (recent reviews, recent social media posts, recent news mentions) are sometimes weighted higher than businesses with no recent activity, though this is less consistent than the other signals.

ChatGPT does not have a formal "ranking algorithm" the way Google does. It uses a large language model to synthesize an answer based on the context it has. The more extractable facts and third-party validation you have, the more likely ChatGPT will cite you.

For example, say a Wilmington landscaper tests ChatGPT and is not cited. He checks his website (vague service descriptions, no pricing, no timeline), his reviews (a few generic Google reviews), and his third-party presence (no news mentions, no Q&A answers, no directory listings beyond Google). The plan writes itself: rewrite the website with clear service descriptions, pricing ranges, and process details; ask recent customers for detailed Google reviews; answer landscaping questions on Houzz; look for a chance to be quoted in a local blog about native plant gardening in Vermont. Extractable facts (pricing, services, process) plus third-party validation (reviews, Houzz answers, a blog mention) are the signals ChatGPT looks for.

Perplexity: the research and shopping specialist

Perplexity is an AI search engine designed specifically for research and discovery. Unlike ChatGPT (which is a general-purpose AI assistant), Perplexity is built to search the web, synthesize information, and provide answers with citations. It launched in late 2022 (around the same time as ChatGPT) but has grown more slowly (around 10 million monthly active users as of late 2024). Perplexity's key differentiator is that it always cites sources (it shows you which websites, articles, or reviews it pulled information from) and it searches the live web in real time (so it is always current, unlike ChatGPT's free version which has a training data cutoff).

Who uses Perplexity?

Perplexity's user base skews toward:

Researchers and students: People who want detailed, cited answers to factual questions. Perplexity is better than ChatGPT for research because it always cites sources (you can verify the information) and it searches the live web (so it is current).

Shoppers and buyers: People researching products, services, or local businesses before making a purchase. Perplexity is strong for "best [product]" or "who should I hire for [service]" queries because it pulls from multiple sources (reviews, blogs, directories) and synthesizes a comparative answer with citations.

Professionals and knowledge workers: People who need accurate, current information for work (lawyers researching case law, consultants researching industry trends, business owners researching competitors or market data). Perplexity is trusted for professional use because it cites sources and you can click through to verify.

People who distrust AI hallucinations: ChatGPT sometimes generates plausible-sounding but incorrect answers (called "hallucinations"). Perplexity reduces this risk by grounding every answer in cited sources. If Perplexity cites a fact, you can click through and verify it. This makes Perplexity more trustworthy for high-stakes questions.

What do people use Perplexity for?

Perplexity is most commonly used for:

Research and fact-checking: "What are the symptoms of X?" or "What is the history of Y?" or "What do experts say about Z?" Perplexity pulls from multiple credible sources and cites them, so you can verify the answer.

Product and service research: "What are the best heat pumps for Vermont?" or "Who are the top-rated accountants in Brattleboro?" Perplexity pulls from reviews, expert blogs, and news articles and synthesizes a comparative answer.

Local business discovery: "Who should I hire for roof repair in Bellows Falls?" or "Where are the best restaurants in Wilmington?" Perplexity pulls from Google reviews, Yelp, TripAdvisor, and local blogs and gives you a list with citations.

Current events and news: "What is happening with [recent event]?" Perplexity searches the live web and pulls from news sites, so it is always current (unlike ChatGPT's free version).

Comparing options: "What is the difference between ChatGPT and Claude?" or "Should I get a heat pump or a furnace?" Perplexity excels at comparative queries because it pulls from multiple perspectives and synthesizes them.

How does Perplexity decide who to recommend?

Perplexity searches the live web in real time and synthesizes answers from the top results it finds. It weights several signals:

Review platforms: Perplexity pulls heavily from Google reviews, Yelp, TripAdvisor, and industry-specific review platforms (Avvo, Houzz, Healthgrades, Angi). If you have detailed reviews on multiple platforms, Perplexity cites you. If you only have reviews on Google, Perplexity may cite you, but competitors with reviews on multiple platforms have an advantage.

Recent content: Perplexity prioritizes recent content (recent blog posts, recent news articles, recent reviews). If your website has not been updated in three years and you have no recent reviews or mentions, Perplexity may skip you in favor of competitors with more recent activity.

Authoritative sources: Perplexity weights news sites and government (.gov) or educational (.edu) sites, plus well-known publications (Wikipedia, industry blogs, major news outlets). If you are mentioned in a local news article or a credible blog, Perplexity often cites that source.

Domain authority and credibility: Perplexity evaluates the credibility of the sources it pulls from. A mention on a well-known site (a local newspaper, a state government site, a national directory) carries more weight than a mention on a random blog.

Structured data and clear formatting: Perplexity can read structured data (schema markup) and clearly formatted content (headings, bullet lists, tables). If your website has well-structured service pages with clear pricing, timelines, and FAQs, Perplexity can extract that information and cite it.

Perplexity is excellent for businesses that have multi-platform presence (reviews on Google and Yelp, a profile on an industry platform, mentions in blogs or news). It is weaker for businesses that only have a website (Perplexity wants third-party validation, not just self-reported claims).

Say a Bellows Falls accountant tests Perplexity and is not cited. He checks the competitors Perplexity recommends. They have reviews on both Google and Yelp (he only has Google). They have profiles on industry platforms (he has none). One was mentioned in a news article (he has no news mentions). His to-do list: claim and complete his Yelp profile, then ask clients for Yelp reviews. After that, answer tax questions on Quora and pitch a guest post about Vermont small business tax tips to a local business blog. Each of those gives Perplexity another independent source to pull from. Multi-platform presence is the key.

Claude: the professional's AI, the thoughtful analyst

Claude (from Anthropic) is the third major AI assistant. It is less well-known than ChatGPT and Perplexity, but it has a loyal user base among professionals, researchers, and people who want more careful, nuanced answers. Claude is known for being more cautious and thoughtful than ChatGPT (it is less likely to hallucinate or make overconfident claims) and for handling long, complex inputs well (it can read and analyze long documents, transcripts, or datasets). Claude is available as a free web app and as a paid subscription (Claude Pro), and it is also embedded in some enterprise tools.

Who uses Claude?

Claude's user base skews heavily toward:

Professionals and knowledge workers: Lawyers, consultants, researchers, analysts, executives. People who need accurate, nuanced answers for high-stakes work. Claude is trusted for professional use because it is careful (it says "I do not know" when it does not have enough information, rather than making something up) and it handles complexity well.

Researchers and academics: People analyzing long documents, academic papers, or datasets. Claude can read and summarize long texts (up to 100,000 tokens, which is roughly 75,000 words), making it useful for research and analysis.

Writers and editors: People who want thoughtful feedback on writing. Claude is known for providing detailed, constructive editing feedback (rather than just rewriting text like ChatGPT often does).

People who want less hype and more substance: Claude's tone is more measured and less "salesy" than ChatGPT. People who are turned off by ChatGPT's over-enthusiastic or overly polished tone often prefer Claude's more neutral, careful style.

What do people use Claude for?

Claude is most commonly used for:

Document analysis and summarization: Uploading a long document (a contract, a research paper, a transcript) and asking Claude for a summary, key points, or answers to questions about it.

Professional research and writing: Drafting reports, white papers, proposals, or analysis with a more formal, careful tone. Claude is good at maintaining a professional voice.

Editing and feedback: Asking Claude to review and critique writing (essays, blog posts, emails). Claude provides thoughtful feedback rather than just rewriting.

Complex problem-solving: Asking Claude to analyze a multi-step problem, weigh tradeoffs, or think through a decision. Claude is good at structured reasoning.

Local business and service recommendations (less common, but growing): Some professionals use Claude to research service providers or local businesses. Claude pulls from web sources and synthesizes recommendations, but this is not as common a use case as it is for ChatGPT or Perplexity.

How does Claude decide who to recommend?

When you ask Claude for local business recommendations, it searches the web (Claude has web search capability in its paid version) and synthesizes an answer based on what it finds. Claude's ranking behavior is less well-documented than ChatGPT or Perplexity, but from testing and observation, Claude weights:

Credibility and authority signals: Claude prioritizes sources it perceives as credible (news sites, professional directories, well-known review platforms, government sites). It is more cautious than ChatGPT (less likely to cite a random blog or a low-quality directory).

Detailed, factual content: Claude rewards depth. If your website has detailed service descriptions, clear pricing, process explanations, and credentials, Claude is more likely to cite you than if your website is vague and promotional.

Third-party validation: Like ChatGPT and Perplexity, Claude weights reviews, mentions, and citations from independent sources. But Claude seems to weight quality over quantity (a few detailed, credible mentions are better than many low-quality mentions).

Recency and consistency: Claude pulls from current sources. If your website has not been updated in years and your most recent review is from 2021, Claude may skip you in favor of competitors with more recent activity.

Claude is excellent for businesses that have strong authority signals (credentials, certifications, professional associations, news mentions) and clear, detailed content. It is less effective for businesses that rely on volume (hundreds of low-quality backlinks or dozens of thin blog posts) because Claude prioritizes quality and credibility over quantity.

Picture a Putney lawyer who tests Claude with queries about Vermont estate planning and family law attorneys and is cited for some queries but not others. The competitors Claude cites instead have more detailed content (comprehensive guides to Vermont estate law, step-by-step process explanations, long FAQ sections). They have also been quoted in credible publications (a statewide news site, a bar association newsletter). She has shorter service pages and no news mentions. Her path: publish in-depth guides (the Vermont estate planning process and family law FAQs) and offer local reporters her expertise when a probate story comes up. She should also add her real credentials to her About page (law school, bar admission, CLE courses, professional associations). Depth of content and authoritative mentions are what Claude rewards.

Why you need to optimize for all three (and what happens if you only optimize for one)

Your customers are not all using the same AI tool. Some default to ChatGPT because it is the most popular and widely known. Some use Perplexity because they want citations and current results. Some use Claude because they are professionals who value careful reasoning. If you optimize for only one tool, you are invisible to the users of the other tools.

Here is what happens if you only show up in one AI tool:

If you only show up in ChatGPT: You are visible to the largest user base (ChatGPT has the most users), but you are invisible to researchers, shoppers, and professionals who use Perplexity or Claude. These are often high-intent users (people actively researching before hiring or buying). You are missing out on some of your best potential leads.

If you only show up in Perplexity: You are visible to researchers and shoppers (high-intent users), but you are invisible to the much larger general audience that uses ChatGPT for local recommendations. You are also invisible to professionals who use Claude. You are capturing a slice of the market, but missing the majority.

If you only show up in Claude: You are visible to a small, highly professional user base, but you are invisible to the 200+ million people who use ChatGPT and the growing number of people who use Perplexity. Claude's user base is the smallest of the three, so optimizing only for Claude means you are missing most of the market.

The businesses that win are the ones that show up in all three. They have the signals each tool prioritizes (detailed reviews for ChatGPT, multi-platform presence for Perplexity, authoritative mentions and deep content for Claude), so they capture leads from all three user bases.

Here is how that can play out. Say a Brattleboro HVAC contractor tests all three tools. ChatGPT cites him fairly often, Perplexity rarely, and Claude not at all. A quick diagnostic shows why:

ChatGPT: He has strong signals (plenty of detailed Google reviews, clear website content with pricing and timelines). ChatGPT cites him consistently.

Perplexity: He has reviews on Google but not on Yelp or any industry platform. He has no third-party mentions (no news, no Q&A answers, no guest posts). Perplexity cites competitors with multi-platform presence.

Claude: His website content is clear but not deep (short service pages, no comprehensive guides, no detailed FAQs). He has no authoritative mentions (no news quotes, no professional association visibility). Claude cites competitors with deeper content and credible mentions.

The fix is to build signals for Perplexity and Claude. For Perplexity: claim and complete his Yelp profile, ask customers for Yelp reviews, answer HVAC questions on Houzz, write a guest post for a Vermont home improvement blog. For Claude: publish in-depth guides (heat pump installation in Vermont, furnace vs. heat pump comparison, HVAC maintenance schedule and costs) and make himself available to local news for heating-cost stories. He should also get listed in the member directory of any trade association he actually belongs to. ChatGPT does not need much more work; the other two are where the leads are being lost.

The core signals that work across all three tools (your foundation)

The good news: the core optimization strategy is the same for all three tools. There are specific tactics that help with one tool more than another (we will cover those next), but the foundation is universal. If you build strong core signals, you will show up in ChatGPT, Perplexity, and Claude.

Here are the core signals that work across all three:

Detailed reviews on Google (and ideally on other platforms too)

All three tools weight reviews heavily. ChatGPT extracts facts from Google reviews. Perplexity pulls from Google, Yelp, and industry platforms. Claude weights reviews as third-party validation. Collect detailed reviews on Google (this is mandatory). If you can, also collect reviews on Yelp and one industry-specific platform (Avvo, Houzz, Healthgrades, Angi, TripAdvisor). The more platforms you have reviews on, the more visible you are across all three AI tools.

Clear, factual website content (services, pricing, timelines, credentials)

All three tools extract facts from your website. Write clear service descriptions (what you do, what is included, who it is for). List pricing (at least ranges or starting prices). List timelines (how long the work takes). Display credentials (licenses, certifications, years in business, professional associations). Use headings, bullet lists, and short paragraphs. Cut vague marketing language. The clearer and more factual your content, the more extractable it is for all three tools.

Third-party mentions (news, blogs, directories, Q&A platforms)

All three tools weight third-party mentions as credibility signals. Get quoted in local news. Write guest posts for regional blogs or industry publications. Answer questions on Q&A platforms (Quora, Reddit, Avvo, Houzz). Get listed on credible directories (Better Business Bureau, industry associations, government contractor lists). Each mention builds your authority and increases the likelihood that all three tools will cite you.

Active, recent presence

All three tools favor businesses with recent activity (recent reviews, recent website updates, recent mentions). If your most recent review is from 2022 and your website has not been updated since 2020, you look inactive. All three tools may skip you in favor of competitors with more recent signals. Post to your Google Business Profile monthly. Collect reviews regularly. Update your website at least annually. Stay active.

Strong Google Business Profile

All three tools pull from Google Business Profile data. Make sure your profile is complete (accurate address, phone, hours, service area, categories, photos). Respond to reviews. Post updates. The stronger your GBP, the more visible you are to all three tools.

If you have these five core signals (detailed reviews, clear website content, third-party mentions, recent activity, strong GBP), you will show up in ChatGPT, Perplexity, and Claude. The differences between the tools matter at the margin, but the foundation is the same.

The specific tactics that help more with Perplexity

Perplexity prioritizes multi-platform presence and recent content more than ChatGPT or Claude. Here are the tactics that give you an edge in Perplexity:

Claim and optimize profiles on multiple platforms

Perplexity pulls from Google, Yelp, TripAdvisor, Houzz, Angi, Healthgrades, Avvo, and other industry-specific platforms. The more platforms you have a complete profile on, the more likely Perplexity will cite you. Claim your profiles (Yelp at minimum, plus one or two industry platforms), complete every section, upload photos, and collect reviews. Perplexity synthesizes from multiple sources, so multi-platform presence gives it more data to pull from.

Publish recent content on your website

Perplexity weights recent content more than ChatGPT or Claude. If your blog has not been updated in two years, Perplexity may skip your site in favor of competitors with recent posts. Publish at least one blog post or guide every quarter. Update your service pages annually. Perplexity wants to see that you are active and current.

Get listed on local directories and "best of" lists

Perplexity pulls from local directories (Yelp, Google, BBB, Chamber of Commerce) and "best of" lists (local publication awards, Yelp's "top 10," regional blogs). Get listed on as many credible local directories as you can. Apply for local awards or "best of" lists. Perplexity synthesizes from these sources when building local recommendations.

Use structured data (schema markup) on your website

Perplexity can read structured data (schema markup) and extract information more easily from well-structured pages. Add LocalBusiness schema to your homepage (business name, address, phone, hours, service area). Add Service schema to your service pages (service name, description, price range). Use FAQ schema for FAQ pages. (For more on schema, see the schema markup post.)

For example, a Townshend contractor who is rarely cited by Perplexity could claim and complete his Yelp and Houzz profiles and collect reviews on both. He could also publish a couple of recent posts and get listed in a local chamber of commerce directory. Finally, he could add LocalBusiness and Service schema to his website. Each item gives Perplexity another current, independent source to pull from. Then test again after a few weeks.

The specific tactics that help more with Claude

Claude prioritizes depth of content and authoritative mentions more than ChatGPT or Perplexity. Here are the tactics that give you an edge in Claude:

Publish comprehensive, in-depth guides (1,500+ words)

Claude rewards depth. Instead of 20 short blog posts (300 to 500 words each), publish 5 to 10 comprehensive guides (1,500 to 2,500 words each). Cover a topic thoroughly (what it is, how it works, what it costs, timelines, FAQs, examples, process details). Claude weights long, detailed content more heavily than short, surface-level content.

Get quoted in credible news or industry publications

Claude prioritizes authoritative sources (news sites, industry publications, .gov and .edu sites). Getting quoted in a local newspaper (Brattleboro Reformer, VTDigger, the Commons) or a credible industry blog carries significant weight with Claude. Pitch local news or write guest posts for industry publications. A podcast interview helps too. Each authoritative mention increases the likelihood Claude will cite you.

Display credentials and professional associations prominently

Claude weights expertise signals (licenses, certifications, professional associations, awards). Add a "Credentials" section to your About page or homepage. List your license numbers, certifications, education, professional memberships, and awards. Link to the licensing board or association directory if possible. Claude looks for verifiable expertise signals, and displaying them prominently helps.

Write in a professional, measured tone

Claude's user base is professional, and Claude itself has a measured, careful tone. If your website content is overly promotional or salesy ("Trusted experts! Best in Vermont! Call now!"), it may feel mismatched to Claude's audience. Write in a clear, professional, factual tone. Explain what you do and how you do it. Then cover what it costs and what customers should expect. Claude rewards substance over hype.

Say a Bellows Falls lawyer is not cited by Claude at all. The Claude-specific tactics give her a clear plan: publish a few comprehensive guides (the Vermont estate planning process, family law FAQs, a Vermont probate guide), make herself available to local reporters covering estate law, add a detailed credentials section to her About page (law school, bar admission, professional associations, CLE courses), and rewrite her homepage in a more professional, less promotional tone. Then test again in a few months. Depth of content and authoritative mentions are the key.

The specific tactics that help more with ChatGPT

ChatGPT has the broadest user base and is the least picky about sources (compared to Perplexity and Claude). The tactics that help most with ChatGPT are the core signals (detailed reviews, clear website content, third-party mentions). But here are a few tactics that give you an edge in ChatGPT:

Collect as many detailed Google reviews as possible

ChatGPT pulls heavily from Google reviews (more than from any other single source). The more detailed Google reviews you have, the more facts ChatGPT can extract. More facts mean it cites you more often. Ask every customer for a Google review. Prompt them to include specifics (service type, timeline, cost, outcome). Aim for 20+ detailed reviews (this is enough to establish a pattern that ChatGPT can extract from).

Answer questions on Quora and Reddit

ChatGPT indexes Quora and Reddit (they are both in ChatGPT's training data, and the paid version can search them in real time). Answering questions related to your industry or location on these platforms increases the likelihood ChatGPT will cite you. Write detailed, helpful answers (300+ words) that include your credentials and location naturally. Link to your website where relevant. Each answer is a third-party signal ChatGPT can extract.

Optimize your Google Business Profile posts

ChatGPT pulls from Google Business Profile data, including posts. Post to your GBP monthly (updates, tips, project highlights, photos). ChatGPT indexes these posts and may extract facts from them (especially if you include specifics like project timelines, pricing, or service details).

Include FAQs on your website

ChatGPT loves FAQ content because it is structured as questions and answers (which matches how users query ChatGPT). Add a comprehensive FAQ page to your website (10 to 20 questions your customers actually ask). Write clear, factual answers. ChatGPT often extracts FAQ answers directly and cites them.

For example, a Londonderry roofer who is cited only occasionally by ChatGPT could ask every recent customer for a detailed Google review, answer roofing questions on Quora and Reddit, post to his Google Business Profile every two weeks (project photos, seasonal tips, pricing updates), and add a FAQ page with fifteen or so real questions (roof repair vs. replacement, cost ranges, timelines, materials, ice dam prevention, insurance claims). Those are exactly the sources ChatGPT pulls from. The volume of detailed signals is what moves the needle.

How to test your visibility in all three tools (the audit)

If you want to know how visible you are in ChatGPT, Perplexity, and Claude, run this audit:

Step one: Write five core queries

Write five queries related to your business and location. Make them natural (the way a real person would ask). Examples:

• "Who are good [your profession] in [your town] Vermont?"
• "What does [your service] cost in [your town]?"
• "Where should I go for [your service] in [region]?"
• "Who should I hire for [your service] in [your county]?"
• "What are the best [your industry] businesses in [your area]?"

Step two: Test each query in ChatGPT, Perplexity, and Claude

Run each query in all three tools. Record if you are cited (yes or no). If yes, note which sources the tool cited (your website, your reviews, a third-party mention). If no, note which competitors were cited and which sources the tool pulled from.

Step three: Calculate your citation rate for each tool

For each tool, calculate your citation rate (number of queries where you were cited / total queries). Example: if you were cited in 3 out of 5 ChatGPT queries, your ChatGPT citation rate is 60%. Do this for all three tools.

Step four: Compare your citation rates

Compare your citation rates across the three tools. Are you strong in one tool but weak in another? If yes, that tells you where to focus. Example: if you are cited in 80% of ChatGPT queries but only 20% of Perplexity queries, you need to build multi-platform presence (Yelp, industry platforms, directories). If you are cited in 60% of ChatGPT queries and 70% of Perplexity queries but 0% of Claude queries, you need to publish deeper content and earn authoritative mentions.

Step five: Analyze the competitors who are cited

For the queries where you were not cited, analyze the competitors who were cited. Visit their websites. Check their review profiles (Google, Yelp, industry platforms). Search for third-party mentions (news articles, blog posts, Q&A answers). What signals do they have that you lack? Reviews on multiple platforms? Deeper content? News mentions? Professional associations? This tells you what to build.

Here is a hypothetical example. A Grafton CPA runs five queries in three tools (15 total tests) and gets something like this:

• ChatGPT: cited in 4 out of 5 queries (80%)
• Perplexity: cited in 2 out of 5 queries (40%)
• Claude: cited in 1 out of 5 queries (20%)

She looks at the gaps. Perplexity cites competitors with Yelp reviews and profiles on industry platforms (she has neither). Claude cites competitors with in-depth guides on their websites and news mentions (she has short service pages and no news mentions). So her plan is: claim Yelp and ask clients for reviews there, claim her profile on an accounting Q&A site, publish a couple of comprehensive guides (Vermont small business taxes, self-employment tax), and pitch a guest post to a local business blog. Then she runs the same 15 tests again and compares.

The audit identifies the gaps; fixing them is what improves visibility across all three tools.

Why "optimizing for AI search" is not three separate strategies

The most important takeaway: you do not need to run three completely different optimization strategies for ChatGPT, Perplexity, and Claude. The core strategy is the same (detailed reviews, clear website content, third-party mentions, credentials, recent activity). The differences between the tools are at the margin (Perplexity wants multi-platform presence, Claude wants depth and authority, ChatGPT wants volume and clarity). If you build the core signals first, you will show up in all three tools. Then you can layer in the tool-specific tactics to strengthen your visibility in the tool where you are weakest.

Here is the priority order:

First 60 days: Build the core signals. Collect 20+ detailed Google reviews. Rewrite your website with clear, factual content (services, pricing, timelines, credentials). Earn 3 to 5 third-party mentions (news quote, guest post, Q&A answers, directory listings). Update your Google Business Profile. This foundation works for all three tools.

Next 30 days: Strengthen your weakest tool. Run the audit above. Identify which tool you are weakest in (ChatGPT, Perplexity, or Claude). Implement the tool-specific tactics for that tool. If you are weak in Perplexity, claim Yelp and an industry platform, collect reviews, publish recent content. If you are weak in Claude, publish in-depth guides and pitch news. If you are weak in ChatGPT, focus on Google reviews and FAQ content.

Ongoing: Maintain and compound. Once you are visible in all three tools, maintain your signals. Collect reviews monthly. Publish new content quarterly. Pitch news or write guest posts twice a year. Answer Q&A questions monthly. Update your credentials and website annually. The businesses that stay visible in all three tools are the ones that make AI optimization a regular part of their marketing routine, not a one-time project.

For a Wilmington retail shop, that could look like this. First 60 days: ask customers for Google reviews, rewrite the website with clear product descriptions and pricing, answer product questions on Reddit, get listed on a Vermont retail directory, then test all three tools. If Claude turns out to be the weakest, the next 30 days go to two in-depth buying guides and a pitch to a Vermont lifestyle blog for a "local businesses to support" roundup. After that, the routine is steady. Aim for a few reviews a month and a guide or post each quarter, plus a twice-yearly pitch to a news site or blog.

What to do if you can only focus on one tool (and which one to choose)

If you are a small business owner with limited time and you can only focus on optimizing for one AI tool, which one should you choose?

Choose ChatGPT if: Your customers are general consumers (homeowners, individual clients, non-professionals). ChatGPT has the largest user base by far, so optimizing for ChatGPT gives you the broadest reach. Focus on Google reviews, clear website content, and FAQ pages.

Choose Perplexity if: Your customers are researchers or shoppers (people who compare options before hiring or buying). Perplexity is strong for "best of" and "who should I hire" queries. Focus on multi-platform reviews (Google and Yelp), recent content, and directory listings.

Choose Claude if: Your customers are professionals or high-end clients (people who value depth and credibility). Claude's user base is smaller, but it skews professional. Focus on in-depth content, authoritative mentions (news, industry publications), and credentials.

For most small businesses serving local consumers, ChatGPT is the best single-tool focus because it has the largest reach. But the smarter strategy is to build the core signals (reviews, website content, third-party mentions) that work across all three tools, then layer in tool-specific tactics as time allows.

How AI tool usage is evolving (and why you need to stay flexible)

AI search is still early. ChatGPT launched in November 2022. Perplexity launched around the same time. Claude became widely available in 2023. The tools are evolving fast, and user behavior is evolving too. Here are the trends to watch:

Perplexity is growing faster than ChatGPT in some niches. Perplexity's user base is growing quickly among researchers, shoppers, and professionals. If current trends continue, Perplexity may rival ChatGPT for research and shopping queries within the next year or two. Businesses that ignore Perplexity now may regret it later.

ChatGPT is adding more real-time search capability. OpenAI is investing heavily in making ChatGPT better at real-time search and local queries. The paid version (ChatGPT Plus) already has web search. If OpenAI makes web search available in the free version, ChatGPT will become even more competitive with Perplexity for current, local information.

Claude is expanding its user base. Anthropic (the company behind Claude) is marketing Claude more aggressively to professionals and enterprises. If Claude gains traction in professional markets, it will become more important for B2B businesses, professional services, and high-end consumer services.

New AI search tools are launching. Google is rolling out AI Overviews (AI-generated answers at the top of Google search results). Microsoft is integrating AI into Bing. Smaller startups are launching niche AI search tools. The landscape is fragmented and competitive. The businesses that build strong core signals (reviews, content, mentions, credentials) will be visible across all these tools, even as new ones emerge.

The bottom line: do not over-optimize for one tool at the expense of the others. Build the core signals that work across all AI tools and test your visibility regularly. Adjust as user behavior and tool capabilities evolve. The businesses that stay flexible and multi-platform will win.

Why testing and iteration matter more than perfection

You do not need to be cited in 100% of queries in all three tools to win. You need to be cited more often than your competitors. If you are cited in 60% of ChatGPT queries, 50% of Perplexity queries, and 40% of Claude queries, and your competitors are cited in 20% or less, you are winning. AI search is not binary (visible or invisible). It is probabilistic (how often are you cited relative to competitors). Every improvement in your signals (one more detailed review, one more third-party mention, one more in-depth guide) raises your citation rate. Each one gives you an edge.

The businesses that win in AI search are the ones that test regularly, identify gaps, fix them, and test again. They do not aim for perfection (being cited 100% of the time). They aim for continuous improvement (being cited more often this quarter than last quarter, being cited more often than competitors). This requires a test-and-iterate mindset, not a set-it-and-forget-it approach.

Picture a Newfane contractor who tests his AI visibility every quarter. In the first quarter he builds core signals (reviews, website rewrite, Google Business Profile). In the second, he works on his weakest tool, say Perplexity (claims Yelp, collects reviews there, answers Houzz questions). In the third, he turns to Claude (publishes two guides, pitches local news). Each quarter's test tells him where to focus next. That loop (test, find the gap, fix it, test again) is how citation rates climb across all three tools.

The bottom line: your customers are using all three, so you need to show up in all three

If you optimize for only one AI tool, you are invisible to the users of the other tools. ChatGPT has the largest user base, but Perplexity is growing fast among researchers and shoppers, and Claude is the choice of professionals. Your customers are using all three, and if you only show up in one, you are losing leads.

The good news: you do not need three completely separate strategies. The core signals (detailed reviews, clear website content, third-party mentions, credentials, recent activity) work across all three tools. Build the core first. Then layer in the tool-specific tactics (multi-platform presence for Perplexity, deep content and authoritative mentions for Claude, Google reviews and FAQ pages for ChatGPT). Test your visibility in all three tools quarterly. Identify your weakest tool and focus on strengthening it. Iterate and improve.

The businesses that show up in ChatGPT, Perplexity, and Claude consistently are the ones that capture leads from all three user bases. The businesses that only show up in one are leaving money on the table. Your customers are searching on all three platforms. Make sure they find you no matter which tool they use.

Every one of these tools reads your website, so the foundation is clear, factual pages. If you need a site, Brattleboro Local builds one for a local business for $499 a year, everything included: six pages written for you from what you tell us (home, about, services, the areas you cover, answers to common questions, and contact), on your own domain with hosting, security and backups kept current. You get a login to change any text, price or photo yourself, and it goes live as soon as it is built, and anything you want changed, we change. Reviews, guides, and news mentions are still up to you. See the details or start your site.