Audience Research
Patterns across many people
We help you understand your audience by what people actually do on your site — not by surveys or demographics. Below we explain the main ideas in plain language.
Audience Research is interpretive: it describes how patterns appear, not what teams should do next.Key terms — in plain language
We use a few words you’ll see everywhere in Beltmar. Here’s what they mean.
What is a cohort?
A cohort is a group of people who behave in a similar way. We don’t group by age, job title, or location — we group by what they do: e.g. “people who keep coming back and read a lot about a few topics” or “people who browse once and drift away.” Each cohort gets a short, readable description so you can see what kind of behavior it represents.
What is a journey?
A journey is one person’s path over time: what they looked at, when they came back, and whether they went deeper or lost interest. Think of it as their story on your site — no grades or scores, just a clear view of how their attention moved. Journeys help you understand individuals; cohorts help you see patterns across many people.
What is an actor?
An actor is simply a person (or anonymous visitor) we’re tracking. We’re not interested in who they are personally — only in how they behave: what they read, how often they return, and whether their engagement is warming up or cooling down. Each actor has a journey.
What are themes?
Themes are the topics or areas of your content — e.g. pricing, onboarding, or product features. We look at which themes show up in someone’s journey or in a cohort. That helps you see what people are focusing on, without ranking or scoring your content.
How we study behavior
We watch how people move around your site over days or weeks — what they read, when they come back, and whether they go deeper or drift away. We then group people who behave similarly into cohorts and describe what we see in plain, narrative language. No scores, no grades, no “conversion” labels.
So: one person’s story is a journey. Many people with similar stories form a cohort. We surface both so you can understand your audience without jargon.
Cohorts
Groups that behave alike
Who deepens, who drifts, who comes back — described in plain language.
Journeys
One person’s path over time
What they looked at, when they returned, and how their interest changed.
Example cohort
“Deep Research”
People with high depth, repeated returns, and focus on topics like pricing. We describe the pattern; we don’t score it.
Like in the product
Daily Brief
Last 24 hours · Sample workspace
New
3
Warming
5
Cooling
2
Themes
2
Who deserves attention · Narrative insights · What changed
Actors
Ava Martinez
24 events · 6 hours ago
Jordan Lee
9 events · 12 hours ago
Sam Chen
14 events · 2 days ago
Cohort
Deep Research Explorers
37 actors · High depth across 3–5 themes, returns after gaps. Pricing, onboarding themes.
Themes & the bigger picture
Themes are the topics or sections people spend time on (e.g. pricing, onboarding). We don’t rank your content — we show how themes show up in people’s journeys: who discovers what first, who goes deeper into which topics, and which themes tend to appear together. That gives you a clearer picture without jargon.
Journey overview is a workspace-level view of how visits typically unfold: return intervals, common paths, and how behavior stacks up over time. Again, we describe; we don’t prescribe.
Themes
Which topics people focus on, and how that appears in discovery vs. deeper research. Perspective, not causation.
Journey overview
How visits usually unfold across your site — return patterns, common paths. We compare; we don’t tell you what to do.
Like in the product
Theme insights
Pricing
research-sustainingRevisited across returns; often with onboarding.
Onboarding
discoveryEarly touchpoint; supports deeper research later.
Journey overview
Observational; no prescriptions.
Interpretation, not prescription
We show what appears to be happening — who’s warming up, who’s cooling off, and what patterns we see — in clear, readable language. We don’t tell you what to do next; we give you the context so you can decide.
In short: we interpret behavior. You take the action.
Does not
- · Forecast, score, rank, or tell you what to do next.
- · Assign outcome probabilities or conversion likelihoods.
- · Replace your judgment with automated recommendations.
Helps you
- · Notice shifts, understand curiosity, decide with awareness.
- · See who’s deepening, who’s returning, and who’s cooling.
- · Read patterns as narrative context, not as action prompts.