How to See Which Competitors ChatGPT Recommends Over You

Key Takeaways

• You can do this manually with no tools: build a list of 15-20 real buyer questions, run them across ChatGPT (logged out, web search on) a few times each, and log who gets named.
• The single biggest mistake is testing your own brand name or generic category terms instead of the actual phrasing a buyer under pressure would type.
• Run each prompt multiple times. AI answers are non-deterministic, one run can miss a competitor that shows up 60% of the time.
• Once you have the list, the useful output isn't "we're behind Competitor X." It's which specific prompts Competitor X wins and you don't, because that tells you exactly what to fix.
• Manual tracking breaks down past about 10-15 prompts across multiple engines run weekly. That's the point to automate, not before.

---

Start with the prompts, not the tool

Most founders skip straight to "what tool checks this" and skip the part that actually matters: the prompt list. A bad prompt list makes any tool useless.

Real buyers don't type your category name into ChatGPT. They type what they're actually trying to solve. Compare:

• Weak prompt: "AI visibility tools" (too generic, pulls a generic answer)
• Real prompt: "how do I know if ChatGPT is recommending my SaaS competitor instead of me"

The second version is closer to what someone under real buying pressure actually types, and it's the version that surfaces a meaningfully different, more specific answer with named tools.

Build 15-20 of these. Pull them from: your own sales calls (what did the prospect say before they found you), support tickets, competitor review site Q&A sections, and Reddit threads in your category (search "[category] reddit" and read what people actually ask each other).

---

Running the check manually

For each prompt, run it in ChatGPT with web search on (the results differ meaningfully from a plain, non-browsing session), and do this at least 3 times per prompt across a few days. Log:

1. Which brands got named, in order mentioned
2. Whether you were named at all
3. What sources ChatGPT cited, if it shows them

Repeat the same prompt set in Perplexity and Google AI Overviews if you have bandwidth. Perplexity shows its sources inline by default, which makes it the easiest engine to audit, if a competitor is named and you're not, the cited source usually tells you exactly why.

Put this in a spreadsheet: one row per prompt, one column per engine, cell value = which brands got named. Patterns show up fast once you have 15+ rows.

---

What to actually do with the list

The output that matters isn't a competitor ranking. It's a per-prompt loss map. You're looking for the specific questions where a named competitor is beating you, because each one points to a specific, fixable gap:

Sort by how often you see the same competitor winning the same type of source, and you'll find one or two fixable patterns instead of a vague "we're losing to Competitor X" feeling. Once you have this list, the natural next step is turning it into a real share-of-voice benchmark instead of a one-off spreadsheet.

---

When to stop doing this by hand

Manual tracking works fine at small scale, roughly 10-15 prompts, checked monthly, one or two engines. It breaks down once you need weekly cadence, more than one engine, or a prompt list past 20-30 questions, because the non-determinism means you need repeated runs to trust the data, and that's a lot of manual querying to sustain.

That's the point where a tool earns its keep. GetIntel runs your real buyer prompt set across ChatGPT, Claude, Perplexity, Gemini, and AI Overviews on a schedule, and gives you the per-prompt competitor breakdown directly, which competitor is beating you, on which exact questions, and what source they're winning from, without you re-running anything by hand.

See your own competitor breakdown across all five engines in one scan.