Why would an AI assistant recommend a competitor over us?

Usually, it’s not a judgment that your business is worse. The assistant may simply have found clearer or more relevant evidence for the competitor—or the user’s prompt may have favored what the competitor offers.

Common reasons include:

• The competitor fits the stated criteria better—for example, price, location, features, availability, or a particular use case.
• Their information is easier to find or compare. Clear, current product or service details help; missing or inconsistent details can make an assistant less confident about you.
• Your site or listings aren’t accessible to the assistant’s search tools. For ChatGPT Search, for example, OpenAI says sites need to allow OAI-Searchbot crawling to be eligible for inclusion. (help.openai.com)
• The answer is based on incomplete or outdated information. Search results and citations can be incomplete, outdated, or incorrect. (help.openai.com)
• For shopping results, the assistant weighs product data and the user’s context. ChatGPT says product results are selected for relevance to intent, using factors that can include structured product metadata and other third-party content; not every matching product will necessarily appear. (help.openai.com)

A useful first step is to test the exact prompt that produced the recommendation, then check what sources it cited and whether your own pages clearly state your strengths, pricing, availability, and who you serve. If you share what “us” is—a product, company, or service—and the prompt where the assistant recommended a competitor, I can help diagnose the likely reason.

AI assistants usually aren't picking a competitor on purpose. The recommendation mostly reflects how well the model knows each brand and how often trusted sources connect that brand to the problem being asked about. The evidence I found is mostly from marketing and SEO vendors rather than peer-reviewed research, so treat it as well-informed practitioner consensus.

Likely reasons

• Training-data footprint. Models learn from a snapshot of web text, and a brand's presence in that snapshot shapes what the model "knows" about it. When someone asks for a recommendation, the model draws on patterns: which brands were discussed most frequently, most authoritatively, and most consistently across many sources.
• Third-party mentions outweigh your own site. LLM visibility isn't primarily about your website. Coverage in publications, review platforms, and forums often matters more than your marketing pages.
• Brand-topic association. The more tightly a brand name sits next to a topic in the training corpus, the higher the probability the model names it for that topic. If a competitor is repeatedly tied to "best tool for X," it gets named for X.
• Entity clarity. Before an assistant can recommend you, it has to correctly identify who you are and which bucket you belong in. If a model can't confidently say what category you belong to, it won't risk recommending you.
• Reviews and sentiment. One vendor's analysis lists five factors: how often a brand appears in training data and retrieved pages, the authority of the sources that mention it, review sentiment, how closely the brand fits the specific query, and structured data that makes the brand legible.
• Live retrieval. Some tools search the web at answer time, so current content matters too. Perplexity and Claude's web search features, and ChatGPT's browsing mode, pull in real-time data on top of their training.
• Incumbency. This training foundation explains why established brands often dominate AI recommendations even when newer alternatives might be superior.

What to do about it

1. Ask the main assistants (ChatGPT, Claude, Gemini, Perplexity) the buying questions your customers ask, such as "best X for Y," and record who gets named and how each one is described. Results vary by platform, so test several.
2. Look at where competitors are mentioned and where you're absent: review sites, comparison articles, analyst coverage, and community forums.
3. Make your category and use cases explicit and consistent across your site, directories, and press. Vendors recommend specific use cases and measurable results over vague positioning.
4. Keep factual details current, since some tools retrieve live pages.

I can't see how assistants currently describe your product, so I can't tell you the specific reason in your case. If you tell me your product, category, and main competitors, I can help you design a prompt test or review your positioning.

It can be incredibly frustrating to see an AI assistant suggest a competitor when you know your business offers a superior product or service. However, it is important to understand that AI recommendations are generally based on "legibility" and data availability, not on the actual quality of your work.

AI models do not "experience" your service or "taste" your craftsmanship. Instead, they function as information-retrieval and synthesis engines that prioritize clarity, trust, and verifiable data.

Here is why your competitor is likely being recommended over you, and how you can change that.

1. The "Confirmation" Trap (The AI’s Memory)

Unlike traditional search engines that crawl the web to discover new information, AI often operates by "confirming" what it already knows.

• Established Authority: Your competitor may have been in the AI’s training data longer. If they have years of press, articles, and forum mentions, the model associates their name with your category by default.
• "Best-Of" Lists: AI models heavily rely on third-party directories, "best of" articles, and industry lists. If your competitor is featured on these lists and you are not, the AI considers them a safer, more "verified" recommendation.

2. Lack of "Legibility"

Even if you are the better choice, the AI may simply be unable to "read" your business effectively.

• Ambiguous Messaging: If your website uses flowery or vague language (e.g., "We create beautiful outdoor spaces") instead of specific, keyword-rich descriptions (e.g., "Paver patios and pool landscaping in [City Name]"), the AI won't know exactly when to recommend you.
• Missing Context: AI needs to see your business name, services, and location clearly linked across different parts of the web. If your competitor uses the exact language your customers type into search engines, the AI will prioritize them because it can match the user’s intent to their content more easily.

3. The "Share of Voice" Gap

AI visibility is often a result of how much third-party "noise" exists about a brand.

• Third-Party Validation: AI trusts what others say about you more than what you say about yourself. If you have fewer reviews, mentions in industry articles, or backlinks from reputable sources, the AI has less "evidence" to support recommending you.
• Missing Metadata: If your website isn't structured for machines—meaning it lacks clear schema, organized service pages, or accessible business details—the AI may skip over it during its synthesis process.

How to Improve Your AI Visibility (GEO)

You can improve your chances of being recommended by shifting your focus to Generative Engine Optimization (GEO):

• Speak Your Customer's Language: Audit your website to ensure it uses the specific, literal terms customers use when searching for your services. Avoid industry jargon that doesn't align with how a typical user asks a question.
• Create "Answer-Focused" Content: Build pages that specifically answer the questions your customers ask (e.g., "What does a patio project cost?" or "What is the process for [X]?"). This makes your content highly quotable for AI models.
• Build a Consistent Digital Footprint: Ensure your business name, services, and location are consistent across all directories, social platforms, and review sites.
• Get Listed: Prioritize being included on independent, reputable industry lists and directories.
• Audit Your "AI Score": You can use tools (or even just perform manual queries on ChatGPT, Perplexity, and Gemini) to see exactly how your brand is being described. If the AI is describing you using a competitor's features, you know you need to update your on-site content to clarify your unique value.

The bottom line: You aren't losing because you're worse at your job; you are losing because you aren't yet "readable" to the AI. By providing the AI with clear, consistent, and factual information, you can transition from being an "invisible" option to a primary recommendation.

An AI assistant recommending a competitor doesn’t necessarily mean the competitor is better. Often, the assistant has clearer or more credible information showing that the competitor fits the user’s specific request.

Common reasons include:

• Closer fit to the prompt: The competitor’s positioning, features, price, location, or use cases match what the user asked for more directly.
• Clearer, more current information: Their pages make details like pricing, availability, integrations, and product capabilities easy to find and verify.
• More independent evidence: Reviews, industry coverage, comparison articles, and customer discussions may give the assistant more confidence in the competitor. AI answers can draw on these sources, not just companies’ own websites. ²⁸
• A visibility or access gap: Your relevant pages may be hard to discover, crawl, or interpret, while the competitor’s are more accessible and directly answer common buyer questions. ⁸
• Different information in different places: Inconsistent or outdated facts about your company can make it harder for an assistant to describe your offering confidently. ⁹

The useful next step is to look at the exact prompt and answer: What need did the user express, which sources did the assistant rely on, and what specific advantage did it attribute to the competitor? Then check whether your own public information clearly and accurately explains how you meet that need—and whether independent sources support it. There isn’t one universal recommendation formula, and different assistants may rely on different sources. ⁸¹⁴