The fastest way to understand what good AI visibility looks like in your market is to check what AI says about your strongest competitor. You don't need special tools — just a browser, 20 minutes, and a notepad.
Run the audit yourself in four prompts
Open ChatGPT or Perplexity (use both if you can — they source differently). Then ask four questions, swapping in your competitor's real business name:
- "What can you tell me about [Competitor Name] in [City]?"
- "Is [Competitor Name] a good [plumber / dentist / bakery]? What do customers say?"
- "What are the hours and location of [Competitor Name]?"
- "Who are the best [plumbers] in [City]?"
That last one is the most important. It tells you whether your competitor gets named unprompted when someone asks a category question — the AI equivalent of ranking first on a map pack. Write down exactly what each model says. You're collecting raw material, not forming conclusions yet.
What to look for in the results
Once you have the AI's responses, check five things:
Are they named at all? In the category question, does their business name appear? If yes, note whether they're first, buried in the middle, or missing entirely.
Is the information accurate? Wrong hours, a closed location, or an outdated service list signal that their AI visibility is poorly maintained — an opening for you.
What sources does Perplexity cite? Perplexity shows footnotes. Those footnotes are the competitor's citation stack: review platforms, local directories, their own website, news mentions. You're looking at the evidence trail AI used to build their profile.
What words does the AI use to describe them? Phrases like "family-owned," "same-day service," or "specializes in emergency repairs" don't come from nowhere. They're pulled from schema markup, review site descriptions, or the competitor's own website copy. That language is a clue about what's actually working.
Does the AI hedge or say "I'm not sure"? Uncertainty means thin data. If a competitor's AI answers are vague or the model declines to answer, they haven't structured their information well — and you can beat them by doing the basics.
Three things competitors get right that you can copy
When you see a competitor generating consistently accurate, detailed AI answers, they're almost always doing one or more of these things:
Structured schema on their website. A properly marked-up LocalBusiness schema block tells AI crawlers the business name, address, phone, hours, and service area in machine-readable format. If the AI knows their hours precisely, schema is probably why. The LocalBusiness schema guide walks through exactly which fields carry the most weight.
A strong directory footprint. Perplexity's citations regularly point to Yelp, Healthgrades, Houzz, Angi, or industry-specific directories. If your competitor is cited from six of these and you're on two, that's the gap. Add the directories they're on, fill out every field, and make sure your name, address, and phone number are identical across all of them — any mismatch weakens the signal.
Specific first-person service descriptions. AI tends to mirror the language businesses use about themselves. A plumber whose website says "we specialize in slab leak detection and repipe projects in the Denver metro" will surface for those exact services when someone asks AI about slab leaks. If your site says "we do all kinds of plumbing" and theirs is specific, their specificity wins.
Where their data actually comes from
Understanding why a competitor ranks well in AI answers means understanding what AI actually reads. Models train on crawlable web content, and retrieval tools like Perplexity pull live data from sources that cite and index local businesses well. Review volume matters, but so do directory listing quality, on-page structured data, and whether any local publications have ever mentioned them.
A dentist in Phoenix who got written up in a "best new practices" roundup two years ago may still be riding that citation in AI answers today. A bakery whose Yelp profile has 400 reviews and a complete menu section will almost always outperform one with 40 reviews and no description. Understanding why AI gets business information wrong covers the full picture — the difference between training-time data and retrieval-time data matters because the fix is different for each.
Turn competitor gaps into your advantage
If a competitor's AI profile is accurate and detailed, your job is to match their infrastructure and differentiate on specifics — service area, specialty, or credentials they don't mention. If their AI profile is wrong or thin, you have a window: build the structure they skipped, and AI will start preferring your cleaner data.
Before copying their homework, check your own starting point. A free scan of your business shows where you already stand on the same questions you just asked about your competitor — no point duplicating their work on topics where yours is already ahead.
Pick your top two or three competitors. Run the four prompts above for each one this week. Write down one specific, concrete thing each of them is doing that you aren't. That short list is your AI visibility work queue.