How to Create a Cohort of High-Intent Visitors in PostHog

Define high intent by behaviour, not a hunch. The people worth grouping are those who viewed pricing, scrolled deep on a key page, or returned more than once — signals that correlate with buying. Build them as a behavioural cohort in PostHog's cohort builder, then use it to segment funnels and focus your attention. A HogQL query sizes the group and checks the definition is catching real intent.

What makes a good high-intent definition

A single action is noisy — anyone can land on pricing by accident. High intent shows up as a combination of signals:

• Viewed pricing — looked at what it costs.
• Returned more than once — came back deliberately.
• Started the signup — the strongest pre-purchase signal.
• Scrolled deep on a key page — actually read it.

Combine two or three so the cohort means something. "Viewed pricing AND visited twice" is far more intentful than either alone.

Build it in the native cohort builder

PostHog's cohort builder is the right tool because a cohort auto-updates as people's behaviour changes:

1. Go to People → Cohorts → New cohort and choose a behavioural cohort.
2. Add a condition: completed event viewed_pricing in the last 30 days.
3. Add a second condition: completed $pageview at least 2 times in the last 30 days.
4. Save it as "High intent — last 30 days."

Now you can pick that cohort as a filter on any insight or funnel, so you can compare how high-intent visitors convert versus everyone else.

Size and sanity-check it with HogQL

Before relying on a cohort, confirm it's neither empty nor everyone. This HogQL counts people who match the core condition:

SELECT
  countDistinct(person_id) AS high_intent_people
FROM events
WHERE event = 'viewed_pricing'
  AND timestamp > now() - INTERVAL 30 DAY

Illustrative sample output:

high_intent_people
1,460

If that number is a sensible slice of your traffic — large enough to act on, small enough to be meaningful — your definition is in the right range. If it's almost everyone, tighten it (add the return-visit condition); if it's a handful, loosen it.

Put the cohort to work

1. Segment your signup funnel by the cohort to see the high-intent conversion rate — it should be far above the site-wide rate, which validates the definition.
2. Review session replays of the cohort who didn't convert — these are your most painful, most fixable losses.
3. Use it for retargeting if you run ads, so spend goes to people already showing interest.

A precise high-intent cohort turns "interested visitors" from a vibe into a group you can measure and act on. If you'd like the reasons high-intent users don't convert found and fixed as a Pull Request, that's what Velyr does.