01
Gather
From your sector, your offer and your core keywords come realistic prompts in the language people actually use with a chatbot.
Prompt research as the basis of every measurement
The wrong questions measure just as cleanly.
A visibility measurement is worth exactly as much as the question set underneath it. We gather the prompts your audience actually uses and rank them by how close they sit to your business.
Why first
The question set decides everything after it.
Every AI visibility metric is a share, and every share has a population. That population is the question set. Build it wrong and you get clean numbers about the wrong thing.
The most common mistake is using keywords as prompts. People type keywords into a search engine and write full sentences to a chatbot, often with a condition and a reason. „Crm software" and „Which CRM suits a ten person agency already on Microsoft 365?" produce entirely different answers.
The second most common mistake is a set made only of questions where you already look good. It reads well in the report and moves nothing.
Method
How we go about it.
02
Cluster
Prompts get grouped by topic. A cluster holding forty variants of one question is one cluster, not forty measurement points.
03
Deduplicate
Wordings that produce the same answer come out. They cost queries and distort the share.
04
Prioritise
Sorted by closeness to your core business. A question that can turn into an enquiry weighs more than a general knowledge question.
05
Enrich
On request with search volume from Sistrix or Ahrefs, so demand sits next to buying proximity.
06
Hand over
As a spreadsheet that runs straight into monitoring. You can keep working with it without us.
Deliverable
What you get.
Scope
prompt research
- Question set prioritised prompts, clustered and deduplicated
- Format spreadsheet, usable in monitoring directly
- Per prompt cluster, intent stage, priority, optionally search volume
- Languages by market; multilingual sets are common
- Reuse yours to keep, including without ongoing support
Prompt research stands on its own. It is also the first step of the audit and the basis for ongoing monitoring.
Limits
What prompt research is not.
It is not keyword research under a new name. There is no search volume for prompts, because nobody publishes what people ask chatbots. Where we add volume it comes from classic search and is an approximation, not a measurement.
It is also no guarantee of completeness. A question set is a reasoned selection. We say how it came about, so you can judge and change that selection later yourself.
And it is not final. Markets shift, wordings change, models get replaced. A set belongs reworked at least once a year, otherwise you end up measuring something nobody asks any more.
Common questions
About prompt research.
How many prompts does a set need?
Twenty at minimum as a rule of thumb, fifty to two hundred is usual. Below twenty the share swings too hard, because a single question already moves it by five points.
Can I not do this myself?
Yes, and for a first impression we recommend it. Writing down twenty questions from your own sales calls costs an hour. The work starts with deduplicating, prioritising and the yearly upkeep.
Where do the prompts come from?
From your offer, from questions your sales people actually get asked, from core keywords and from what chatbots themselves suggest as follow-ups. Not from an off the shelf list.
Is there search volume for prompts?
No. No provider publishes what people ask chatbots. Where we add volume it comes from classic search and is an approximation.
What does prioritising by closeness to the business mean?
A question that can turn into an enquiry sits at the top. A general knowledge question in the same field sits lower, even when it gets asked more often.
Does each language need its own set?
Yes, and not a translated one. The same intent is phrased differently in another language, and the answers get built from different sources.
How often does a set need reworking?
Annually at minimum, more often in fast moving markets. Anyone who never touches a set ends up diligently measuring something nobody asks any more.
Can I keep the set if we stop working together?
Yes. You get it as a spreadsheet and can keep measuring without us. That is deliberate.
Does this suit very small niches?
Those especially. In narrow niches the set is smaller, every single question sits closer to the business, and results get unambiguous faster.
How does this relate to the audit?
Prompt research is the first step of every audit. As a standalone service it suits anyone measuring themselves or putting an existing monitoring setup on a better basis.
Do you measure straight away as well?
On request. The research ends with the question set. What comes after sits under audit and monitoring.
What does prompt research cost?
Price follows the number of languages and the breadth of your offer. The ranges sit on the pricing page, and if your case does not fit them we give you a number in the first call.
Difference
How others do this, and how we do.
Not a judgement on individual agencies, but a description of two approaches. The right column is checkable, because this website itself is built that way.
| Difference | Usual agency approach | agenticsearch.at |
|---|---|---|
| Starting point | A proposal after a first call | A measurement with a date, before anything gets recommended |
| Basis | Keywords from classic search | A question set of real prompts, clustered and prioritised |
| Proof of success | Rankings and a visibility index | Mentions per question and system, with the raw data |
| Reporting | A monthly dashboard with curves | The same measurement repeated, change called only after three runs |
| When unsure | Try it and see what happens | We say what is not evidenced, and write it on the page |
| Tooling | Agency licences, access ends with the contract | The question set and the data are yours, afterwards too |
| Scope | As many channels as possible from one supplier | A capped number of concurrent projects, otherwise we turn it down |
Your guides
You play the lead.
You know your business, your customers and your offer better than any agency does. What is usually missing is solid numbers on how AI systems talk about it, and an order in which to tackle that.
Who is behind this
two people
- Florian Hieß measurement, technology, structure
- Romana Hieß day to day support and implementation
- Operated by Digital Wings GmbH, founded 2010
- Client projects 100+
The number of concurrent projects is capped. When a piece of work needs more capacity than exists, we turn it down rather than take it on.
01
You know where you stand
It starts with a measurement, not a proposal. You see in black and white which AI answers name your company and which do not, with a date and a question set.
02
You know what comes next
Instead of a list of everything, you get an order. Every step has a result you can verify yourself, including without us.
03
You keep the decisions
Your offer and your positioning stay yours. We tell you what stands in the way of visibility and take on the technical part of it.
Next step
Let us find out where you stand in AI answers.
One conversation, 30 minutes, no sales pressure. You leave with an honest read on whether this is worth the effort for you.
