What a buyer question is, and how to pick yours

Track the questions a buyer types before they know your name, in the buyer's own words, with the qualifiers a buyer would include. Those are the questions where being named changes the outcome. Questions containing your brand name measure your prompt, not your visibility.

What is actually happening underneath

A buying journey produces four kinds of question, and a set missing any of them has a blind spot.

Category questions. Best inventory software for a small manufacturer in India. These are the highest value and the hardest.

Problem questions. How do I stop losing track of stock across two warehouses. The buyer has not decided a category yet. Often the easiest place for a smaller company to be named, because fewer vendors write for them.

Comparison and alternative questions. Alternatives to X. X or Y for a mid sized firm. Your competitor's name does the searching, and you can be the answer without ever ranking for your own category.

Choosing and price questions. What should I expect to pay. What should I check before buying. These reveal whether the assistants have your pricing and positioning right.

The qualifiers matter as much as the category. India, the city, the company size, the industry, the budget, the regulation. A question without them pulls global sources and global names. A question with them pulls the sources your buyer would actually land on.

The six qualifiers to vary, and where to get the wording

A set of twenty questions is really a small grid. Vary these six and the grid builds itself.

One, place. India, a state, a city. Most Indian buying questions carry one.

Two, company size. A single owner, a ten person firm, a mid sized company, a large enterprise.

Three, industry. Manufacturing, retail, professional services, whatever you actually serve.

Four, budget. Free, cheap, under a monthly figure, no constraint stated.

Five, regulation or standard. The compliance your buyer must meet, where it applies to what you sell.

Six, urgency or stage. Starting out, switching from another tool, scaling up.

Four places to get the real wording, in order of usefulness. Ask two salespeople for the last ten questions a prospect asked, verbatim, without cleaning up the grammar. Read your support inbox. Read your search console queries. Read the question titles on the forums where your buyers already ask. Your own marketing copy is the worst source, because it is written in your vocabulary rather than theirs.

A worked example

On 17 September 2026, in a Claude run on the category of clawlaw.in, our sister company, one group of questions behaved differently from every other group in the set. They were the questions about how to look a case up.

On those, official court portals took every position above any commercial product. The assistant said plainly that no commercial product should rank above the official portal for a question about using that portal.

That is a question type telling you it cannot be won, and it is worth knowing before you fund a page for it. Nothing you publish makes you the right answer to how do I use this portal, when the portal itself is right there.

It also points at the questions next door, which are winnable. The buyer who asks how to use the portal usually has a problem the portal does not solve: doing the same job for many companies, keeping a record of it, or checking more than one court at once. Those are problem questions, they carry the qualifiers your buyers use, and they are the rows that should replace the unwinnable ones in your set.

What it means for a business

Use the buyer's vocabulary even when it is less precise than yours. If they say company case check and you say litigation due diligence, track the first and write the page in both. Your internal naming is invisible to the question.

Spread the set across everything you sell into. If you serve three cities and two customer sizes, a set that only asks about one city and one size will look better than reality and hide where you are absent. Our own tracked sets start at 100 questions for this reason, spread across the zones the business actually sells into.

Keep brand questions in a separate sheet with a separate total. They are worth asking, for a different purpose, and they must never be counted in the visibility number.

What a question set cannot tell you

Five limits, and the first is the one that gets overclaimed in reports.

It is not search volume. You do not know how many buyers ask these questions or in what words, and there is no reliable public source for the wording people use inside assistants. A set is a sample of plausible questions, not a demand estimate.

It cannot cover everything. Any set omits questions, and the ones you omit are usually the ones you did not think you could win.

A comfortable set flatters you. If the number looks good immediately, suspect the set before celebrating.

It ages. Vocabulary changes, products change, new competitors appear. Review the set every few months and add rather than replace.

It does not measure intent to buy. Being named in a problem question is early in the journey. Useful, and not the same as being named where the decision is made.

Common questions

How many questions should I start with?

Twenty to thirty, spread across the four kinds and your real zones. Ten is a workable first hour. Five is five answers rather than a rate.

Should I include my competitor's name in a question?

Yes, in the alternatives questions, because buyers genuinely type those and you can be named there without ranking for your own category. Keep them in the main set. It is your own brand name that has to stay out.

English or a local language?

Ask in the language your buyers use. If they mix English with a local language, ask both ways and log them as separate rows, because the sources retrieved can be completely different.

How do I know whether a question is realistic?

Read it aloud. If it sounds like marketing copy or contains a phrase only your team uses, rewrite it. The test is whether a buyer who has not decided anything yet would type it.

Should the set include questions I expect to lose?

Yes, and that is how you know the set is honest. A set built only from your strengths produces a number that feels good and cannot guide a decision.

Can I use keyword tools to build the set?

They help with vocabulary and they are not the same thing. Search keywords are short and typed for a results page. Assistant questions are longer, more conversational and carry qualifiers. Use the tools for words, and your sales team for sentences.

What to do first

• Ask two salespeople for the last ten questions a prospect asked them, verbatim. Do not clean up the wording.
• Read your own support inbox and your search console queries for the same thing.
• Write twenty to thirty questions, covering all four kinds, with real qualifiers.
• Delete any that contain your brand name. Move them to the separate sheet.
• Read each one aloud and ask whether a buyer would really type it. If it sounds like marketing, rewrite it.
• Save the set in a file with a date, and treat it as fixed until you deliberately add to it.

The set will feel uncomfortable when it is right, because it will include the questions you expect to lose.