Most brands have never seen the questions their customers put to AI. They know the questions they would like to be asked. They do not know what a buyer in their market actually types into ChatGPT at 11pm, whether their name comes up, or which competitor gets recommended instead. That gap is the reason AI visibility is so hard to manage: you cannot fix what you have never looked at.
As of this week, every brand can look, free. Enter your website on the HiBot homepage and in about 30 seconds you get your AI Visibility Question Set: the real questions customers in your market ask AI about you and your competitors. No account, no card, no call.
What You Get
HiBot reads your homepage, works out what you sell, who you sell to, and who you compete with, and writes the questions a buyer in that market would ask an AI assistant. Ten appear on the page right away, spread across all six ANSWER categories so you see the full shape of the set, not the first ten in a list. Enter your email and the other 40 open in place. The whole set of 50 is also sent to you, grouped by category, so you can forward it or come back to it later.
If the inferred profile is off, a competitor you never think about or a market you do not sell into, you can fix it and regenerate the set once. The questions are specific to your brand, your rivals and your market. They are not a template with your name pasted in.
Why These Questions Are Worth Having
A question set is a map of the moments where an AI answer can win or lose you a sale. Read your 50 and you will notice things. Some questions never mention your brand at all, because that is how customers actually shop: "What is the best KPI database for a finance team?" has no brand in it, and whether AI names you in the answer is the whole game. Other questions name a competitor next to you, and you may not have known that AI treats that company as your peer. A few ask about your price, your features or your reputation in words you would never use yourself.
That is the value before a single answer has been collected. The set tells you where to look. Run even a handful of the questions in an AI assistant and you will learn more about your AI visibility in ten minutes than most marketing teams know today.
Built on the ANSWER Framework
Every set follows the ANSWER framework, six things AI decides about a brand across the buying journey: whether it names you unprompted (Appearance), whether it recommends you for a specific need (Nomination), how you fare head to head (Showdown), whether it gets your price and value right (Worth), whether it describes what you do accurately (Expertise), and whether it reflects your standing and reviews fairly (Reputation). The categories map to the stages of the customer decision journey, which is why we built ANSWER on McKinsey's model rather than on a keyword list.
The allocation across categories is the same one our paid audits use, which is what makes the free set more than a teaser. A Pulse audit runs 25 questions with a fixed shape across the six categories, and a Panorama runs 50. Your free set is that 50. If you buy an audit, these exact questions are the ones we run, and you can edit, add or swap any of them before we start.
For the methodology in full, the ANSWER whitepaper is free, and the resources hub collects our category-by-category guides, including worked prompt lists such as the 10-prompt Appearance self-check and 12 Showdown prompts for competitor comparisons.
Run the Audit Yourself
Once you have your set, you can run the audit on your own. The method matters more than the tooling. Use a fresh, logged-out session for every question so the engine has no memory of you or your earlier prompts. Ask each question on each engine you care about, word for word. Record the answer verbatim. Then score every answer against the 0 to 4 rubric described in How an ANSWER Audit Is Scored, and roll the scores up by category. How an ANSWER Audit Is Run walks through the capture conditions in detail.
Be honest with yourself about the arithmetic. The full set across five engines with two runs each is 500 prompts, roughly 38 hours of collecting and reading before any scoring. The Pulse shape, 25 questions on three engines with one run each, is 75 prompts and about 12 hours. Many teams run five or ten of their questions by hand, learn something uncomfortable, and then decide how far to take it.
Or Have HiBot Run It for You
The other path is to hand the set to us. A HiBot specialist runs every question, logged out and in a fresh session, on each engine in your plan, records what comes back, and scores each answer against the rubric. You get a report with your HiBot Score, a competitive analysis, the claims AI states about you as fact, prioritized recommendations, a 90-day roadmap and the full audit log. Inside the New HiBot Audit Report shows what that looks like, and there is a complete sample report you can read end to end.
- HiBot Pulse, $199. 25 of your questions on ChatGPT, Perplexity, Google, 1 run each. The focused read.
- HiBot Panorama, $499. All 50 questions on five engines, everything in Pulse + Claude and Grok, 2 runs each and averaged, with Share of Voice against your competitors. The deep read.
Both plans can run on a recurring monthly or quarterly schedule at 20 percent off, which is how you track your AI visibility over time on a locked question set. Reports arrive within one business day for Pulse and two for Panorama, every audit carries a 30-day money-back guarantee, and buying from your question set page takes one click because the questions are already yours.
Why We Give the Questions Away
Because the questions are not the hard part. Generating them is automated. The hard part, and the reason a HiBot audit is worth paying for, is the human-run capture: hundreds of prompts asked under neutral, logged-out conditions, recorded faithfully, and scored by a specialist who reads every answer. Giving away the questions shows you exactly what that labor consists of. Whether you then do it yourself or ask us to, you are deciding with the map in your hand.
Common Questions
Is the question set really free?
Yes. Ten questions appear with no account and no email. The remaining 40 open when you enter an email, and the full set is sent to that address. There is no card and no obligation.
How are the questions generated?
HiBot reads your homepage, infers your category, market and competitors, and writes questions a buyer in that market would ask an AI assistant, following the ANSWER framework's allocation across its six categories. You can correct the inferred profile and regenerate once.
Does HiBot score my website or crawl my site?
No. HiBot fetches your homepage once, only to prefill the brand details you would otherwise type. The audit measures AI answer engines, not your website, and the fetched text is never used in scoring or in a report.
If I buy an audit, does it use the same questions?
Yes. A Pulse audit runs 25 of them and a Panorama runs all 50, and you can edit, add or swap questions before the audit starts. Nothing is regenerated behind your back.
Ready to see yours? Enter your website on our homepage. It takes about 30 seconds.
