Friction Analysis

Friction Analysis surfaces the pages and steps where users most often stall, abandon, or need guidance — so your team can fix the underlying UX instead of just re-guiding around the problem.

Friction score

Each page in your product gets a friction score derived from aggregated session data. The score rises with stall events, flow abandons, and repeated re-entries on the same step.

Signals that raise the friction score:

• Users idling more than 18 seconds without progressing (stall events)
• Flow abandonment at a particular step more than once per session
• Users re-entering the same flow across multiple visits
• Sting bookmarks placed by support agents on specific targets

Friction dashboard

The Friction Analysis view shows a ranked list of pages by friction score, with drill-downs for each page:

• Stall count — total stall events recorded
• Abandon rate — percentage of flows that did not complete
• Completion rate — percentage that reached end_flow
• Avg dwell time — how long users typically spend on the page
• Most-stalled step — the target where stalls concentrate

Acting on friction data

Use these questions to decide whether to fix the product UX or improve guidance:

• Is the target ambiguous? If many users stall at the same input, fixing the label or placeholder copy is better than adding a bee step.
• Is the step order wrong? If users try actions before the form is ready, the page flow may need reordering.
• Is the task genuinely complex? Some tasks benefit from guided flows. High friction on a configuration page often means a flow is warranted.

Honey and sting markers