How AI Chatbots Decide Who to Recommend (And What to Charge to Fix It)

AI chatbot recommendation is now a real acquisition channel, and most businesses have no idea how it works. When someone asks ChatGPT 'what's the best accounting firm for small businesses in Denver' or asks Perplexity to recommend a local HVAC contractor, the AI doesn't run a search and hand back a list of paid ads. It synthesizes what it already knows — or can retrieve — about businesses in that category, applies a set of implicit criteria, and picks. This post explains what those criteria are, why most businesses fail them by default, and what a well-run agency can charge to fix it.

The Short Version of How AI Recommendations Work

Each major AI platform handles recommendations slightly differently, but the underlying logic is consistent. ChatGPT (especially with browsing enabled) pulls from indexed web content and its training data, weighing sources it considers authoritative. Perplexity runs live web searches and surfaces businesses whose information appears clearly structured and frequently cited. Claude tends to be conservative — it's more likely to say 'I can't recommend a specific business' unless the evidence is strong. Gemini integrates Google's knowledge graph, which means businesses with strong Google Business Profiles and structured data have a real edge there.

The common thread: all of them are trying to find businesses they can confidently describe. If an AI can't quickly answer 'what does this business do, who do they serve, where are they, and why are they credible,' it won't recommend them. Not out of bias — out of uncertainty. AIs don't recommend what they can't explain.

What AI Systems Are Actually Evaluating

• —Clarity of service description: Is it immediately obvious what the business does and for whom? Vague taglines like 'solutions for modern challenges' don't help an AI explain you to a user.
• —Structured data presence: Schema markup (LocalBusiness, Service, FAQ, Review) tells AI systems what your data means, not just what it says.
• —Third-party mentions and citations: Perplexity in particular looks for businesses mentioned across multiple sources. One good website isn't enough.
• —Review signals: Not just star ratings — the text of reviews matters. If reviews use category-specific language ('fixed our commercial HVAC system fast'), that content informs AI responses.
• —Entity consistency: Is the business name, address, phone, and description consistent across your website, Google, Yelp, industry directories, and news mentions? Inconsistency reads as noise.
• —Authoritative content: Blog posts, FAQ pages, and how-to guides that answer real questions put you in the position of being the source AI cites — not just the business AI describes.

The Counterintuitive Part

High SEO rankings do not translate directly to AI recommendations. This surprises people. A business can rank #1 on Google for a competitive keyword and still be invisible to ChatGPT. Why? Because traditional SEO optimizes for crawlers and ranking algorithms. AI recommendation optimizes for comprehension. A page stuffed with keywords but light on clear, factual, well-structured content about what the business actually does will rank fine on Google and get ignored by Gemini. The inverse is also true — a well-structured business with modest SEO rankings can show up consistently in AI recommendations because the information is clean, credible, and easy to synthesize.

What Agencies Can Actually Offer — and What to Charge

This is where it gets practical for agency owners. AI visibility optimization (AEO — Answer Engine Optimization) is a real, billable service. It is not SEO repackaged with a new acronym. The deliverables are distinct.

• —AI Visibility Audit ($500–$1,500 one-time): Score the client's current AI visibility across ChatGPT, Perplexity, Claude, and Gemini. Identify what's missing — schema, entity consistency, content gaps, citation gaps. Tools like Zygur automate the scoring so you're not doing this by hand.
• —Foundation Fix Package ($1,500–$4,000 one-time): Implement schema markup, rewrite key service pages for AI clarity, create or update FAQ content, fix entity inconsistencies across directories. This is the work that moves the score.
• —Ongoing AEO Monitoring Retainer ($500–$1,500/month): AI systems update. New sources get cited. Competitors improve. Monthly monitoring tracks whether the client is still showing up in AI responses for their target queries, and flags when fixes are needed.
• —Competitive Gap Analysis (add-on, $750–$2,000): Compare the client's AI visibility to their top three competitors. Show exactly why Perplexity recommends a competitor and not them. This sells fast because it's concrete.

The margin on AEO work is solid. Most of the audit and scoring is automatable. The fixes are content and technical work your team already knows how to do. The retainer is defensible because AI recommendation is not a one-time fix — it's an ongoing position that requires maintenance.

How to Start the Conversation With a Client

Don't open with 'AI SEO.' Open with a demonstration. Pull up ChatGPT or Perplexity in front of the client, type 'recommend a [their category] in [their city],' and show them the results. If they're not there, the conversation sells itself. If they are there, show them how thin the description is compared to a competitor who's been optimized. Either way, you have a problem to solve.

Frequently Asked Questions

Does AI visibility optimization work differently for local businesses versus national ones?

Yes. Local businesses benefit heavily from Gemini's integration with Google's knowledge graph, so a strong Google Business Profile is foundational. Perplexity tends to surface local businesses when location-specific queries are used. National or service-area businesses need to focus more on content authority, third-party citations, and structured data that clearly defines their service scope.

How long does it take to see results after optimizing for AI visibility?

Faster than traditional SEO, typically. Perplexity can surface a business within days of a site update because it runs live searches. ChatGPT with browsing enabled responds similarly. Claude and Gemini may take longer because they depend more on training data cycles and knowledge graph updates. Most clients see measurable changes in AI responses within two to six weeks of completing foundational fixes.

What's the difference between AEO and traditional SEO?

Traditional SEO is about ranking — getting your page to appear in a list of results. AEO is about being selected and described accurately by an AI that's trying to answer a user's question directly. The techniques overlap in some areas (structured data, content quality) but diverge significantly in others. Keyword density is mostly irrelevant for AEO. Entity clarity and citability are everything.

Can a business hurt its AI visibility by doing something wrong?

Yes. Inconsistent business information across directories is a common one — it creates ambiguity that AI systems avoid. Overly promotional website copy that makes factual claims hard to extract also hurts. And businesses that have removed structured data or switched domains without proper redirects often drop out of AI recommendations entirely. The bar for AI recommendation is clarity and credibility, and anything that muddies either will cost you.

Do I need to optimize separately for each AI platform?

Not entirely. The foundational work — clean structured data, clear service descriptions, entity consistency, strong review signals — helps across all platforms. Platform-specific nuances matter at the margin: Gemini rewards Google ecosystem presence, Perplexity rewards being cited across multiple live sources, Claude rewards conservative factual accuracy over marketing claims. A solid foundation handles 80% of the work; platform tuning handles the rest.