Platform · Question Analysis
Decide what to measure, before measuring
BIKMA builds the prompt perimeter on what your market actually says, not on an internal panel.
The brands we measure every day
The two levels
Choose how deep the perimeter goes
Level 1 is included in every plan. Level 2 adds the voice of the market.
Included in every plan
Methodological prompt generation
A sampling structure derived from the literature on prompting and information seeking, adapted to each category and need.
- Eleven need families, weighted by vertical.
- Available from day one, free trial included.
- Editable and versioned: every change stays on the record.
| Availability | Every plan, free trial included |
|---|---|
| Where the questions come from | Scientific literature on prompting and information seeking |
| What it guarantees | A coherent set to start measuring right away |
| What it does not do | It does not surface the emerging themes of your market |
| Timing | Immediate |
What sets us apart
Market analysis of demand
We survey the public conversation in your sector: the perimeter is not what you assume, it is what the market says.
- Questions drawn from forums, communities, publishers and social.
- Every prompt weighted with the Prompt Probability Index.
- Methodology published and validated with the University of Pisa.
| Availability | Dedicated service, on plans with market analysis |
|---|---|
| Where the questions come from | Real public conversation in the sector |
| What it guarantees | Questions the market actually asks |
| What it does not do | It is not instant: it requires sector analysis |
| Timing | Two to four weeks |
The metrics
The demand perimeter, built and weighted
From the public conversation in your market to the prompt set we measure every day: five readings that turn a list of questions into a plan.
-
Branded and non-branded split
The branded set measures recognition, the non-branded one discovery. They stay separate because they answer different questions.
Perimeter prompts mock-up6.437 prompt- Non-branded 91%
- Branded 9%
-
Popularity by cluster
What counts is not how many questions a cluster holds, but how much real demand they carry: that is the basis for priority.
Popularity by cluster mock-up
-
Prompt Probability Index
Every question has a probability of being typed in that market and that period: priorities can be read before you spend.
Probability of being typed prompt selezionatoHigh- Low
- Medium
- High
-
Eleven need families
In CPG the usage occasion matters, in OTC safety, in B2B comparison: the demand profile changes by vertical.
Weight by family CPG vertical
-
The monitored prompts
Monitoring is not a flat list: every prompt carries its cluster, estimated popularity and coverage status.
Prompt · popularity mock-up- best supplement for athletes 5,6K High
- how to choose a supplement 2,9K Medium
- cheap supplements online 886 Low
The questions we measure are not invented: they come from the scientific literature and from a model validated with the University of Pisa.
Frequently asked questions
Taken verbatim from the monitored prompts. FAQPage schema.
How are the prompts to monitor chosen?
On two levels: a structure derived from the literature on prompting, and the extraction of questions from real public conversation in the sector, weighted by how likely they are to be typed.
How can you know what people ask AI about a brand?
There is no public log of those conversations. It is estimated from the real public conversation in the domain, with a declared margin of error.
Is the market analysis included in the free trial?
No: the free trial includes the methodological generation. Market analysis is a dedicated service.
How long does it take?
Two to four weeks, depending on how broad the market is.
Want to see your brand perimeter?
We build the Question Analysis on your category and show you the map of demand on your own data.
