Use case
Audience discovery
Find the segments that actually exist in a category, not the ones on a slide.
The problem
Why this is hard today.
Most segmentation decks are a year old and were never built from anything anyone said. Meanwhile the segments that drive your ratings, your churn and your support load are sitting in public text nobody has clustered.
How userken does it
Four steps.
- Choose a category, a competitor, or a filtered slice such as one platform or one rating band.
- userken embeds the reviews, clusters them, and names each cluster from the language inside it, with a weight for how large the segment is.
- Read each persona's keywords, themes, rating range and a real representative quote.
- Save the audience, then ask it questions or interview it in a focus group.
build_audience
Define an audience by filters and build personas for it.
get_all_personas
Read the personas and weights for an audience.
semantic_search_reviews
Find the concept, not the keyword, across the corpus.
Tool names are the MCP tools and in-app templates. The same engines power both.
Illustration
What a run looks like.
Illustrative example, not a result we are reporting
Illustration: mapping the news category
- Audience: every indexed news app, all ratings, last 24 months.
- What comes back: persona clusters with weights, each with the themes and the rating range that defined it.
- What you would do next: pick the two segments you did not know you had and run a concept test on each.
Related
Next door to this.
Run a audience discovery on your category.
Accounts are provisioned by hand during early access. Tell us the question and the category and we will set you up.