From zero to a tracked brand takes about fifteen minutes. This page walks the whole path: free audit, signup, brand setup, fact verification, first sweep, and how to read what comes back.
1. Run the free audit
Start at /audit — no account needed. Enter your brand name (and optionally your domain) and Brandflare runs a compact live check:
- It researches your brand: category, two to three direct competitors, and five high-volume buyer questions where you should ideally appear. The prompts never name your brand — they're the generic questions a real customer asks an AI assistant.
- Each of the four engines — ChatGPT, Gemini, Perplexity, and Claude — answers two of those prompts, about eight generations in total, and every answer is judged.
- You get a results page showing, per engine: whether you were mentioned, your rank when you were, which competitors appeared, and the worst false claim the judge caught — with an excerpt of the actual answer.
The audit is free, takes around five minutes, produces a shareable results link, and is limited to one per day. It's a snapshot, not a trend — that's what tracking is for.
2. Sign up
Create an account at /sign-up. Every account starts with a 14-day free trial at Pro depth — daily sweeps, 40 prompts, 5 competitors, accuracy alerts — with no credit card required. Plan details are in plans and limits.
3. Add or claim your brand
Two ways to get a brand into your portfolio:
- Add it fresh. Enter your brand and Brandflare sets it up: category, aliases and product names (so mentions under any name are caught), competitors, and a generated prompt panel of real buyer questions.
- Clone it from the catalog. If your brand is already openly tracked on Explore, you can clone it straight into your private portfolio and start from its existing setup.
Review the generated setup before moving on. The prompt panel is your yardstick — it stays fixed so trend lines mean something — and the competitor list determines who you're measured against. Both are worth two minutes of scrutiny now; see prompts and the prompt matrix for what makes a good panel.
4. Verify your fact sheet
Your fact sheet is the ground truth Brandflare checks AI answers against: pricing, product names, what's current and what's discontinued, founding facts. Brandflare drafts facts automatically when the brand is set up; your job is to review them and verify the ones that are correct — and fix or remove anything that isn't.
This step matters more than it looks. Every claim an engine makes about your brand is compared against these facts, and affirmative contradictions become severity-graded flags. A wrong fact in your sheet means wrong flags; a thin sheet means missed hallucinations. Details in accuracy monitoring and flags.
5. Run your first sweep
A sweep runs your entire prompt panel against all four engines and judges every answer. On the trial (and Pro and Agency plans) sweeps run daily; on Starter, weekly. You can also trigger a sweep on demand from the brand page rather than waiting for the schedule — worth doing right after setup so your dashboard has data immediately. The full pipeline is documented in how sweeps work.
6. Read the dashboard
After your first sweep completes, the brand dashboard fills in:
- AI Visibility Score — your average mention rate across engines, 0–100, over a trailing window of roughly a week. One sweep gives you a starting point; the trend takes a few sweeps to become meaningful. Methodology: how the score is computed.
- The prompt matrix — per-prompt, per-engine results showing exactly which questions you win and where you're absent, with rank movement between sweeps.
- Open flags — false claims the judge caught, graded by severity, each linked to the verbatim answer and the fact it contradicts.
- Competitor metrics — share of voice and head-to-head mention rates from the same answers.
What to do in week one
Resist the urge to react to any single answer — engines are non-deterministic, and one run is noise. Instead: verify your fact sheet completely, confirm your competitor list is the set you actually care about, and let a few sweeps accumulate. By the end of the trial you'll have a real baseline, a list of prompts you're losing, and any accuracy problems surfaced — which is exactly the information you need to decide what to fix first.