Experience

Experience brings issues, heatmaps, journeys, returning visitors, and recordings into one investigation workspace. It connects an observed behavior to aggregate impact and, when a recording exists, the exact moment worth watching.

leadmaps detects and groups the behavior before any optional AI summary runs. AI can explain coarse aggregate facts, but it does not decide whether an issue exists, how important it is, or whether an outcome difference is real.

Set it up

1. Use consent mode and connect your consent banner to the leadmaps consent API. Experience never starts in anonymous privacy mode.
2. Open Experience > Settings, choose the site, and enable Experience. Existing sites start disabled so their collection does not change unexpectedly. New consent-mode sites can start enabled under the workspace policy. Only workspace owners and admins can save changes; other members can inspect the effective state.
3. Leave the browser SDK option omitted to follow the remote policy, set experience: true explicitly, or add local restrictions such as blocked selectors and disabled form progression. See the Experience SDK module.
4. Visit the site with consent granted, then use Experience in the dashboard to inspect issues and heatmaps.

The site policy remains authoritative. A page can reduce capture, but it cannot raise sampling or enable a category that the server disabled. Active pages refresh the policy about once a minute and when they return to the foreground. If a refresh fails, capture stops until a later policy request succeeds.

Customer settings and rollout state

The Settings page separates your saved preference from the effective capture state. You can:

• enable or disable Experience for one site;
• enable safe form-progression counters;
• choose eligible adaptive recording reasons on Pro and Business; and
• exclude sensitive page regions with CSS selectors.

Sampling, detection rules, plan eligibility, consent requirements, and rollout safety limits are read-only. A saved setting may therefore show as inactive while a privacy, plan, or rollout condition blocks capture. The page explains which condition applies instead of treating a successful save as proof that capture started.

What each area does

• Issues groups recurring friction into a prioritized list. Open an issue to see its trend, affected-session range, outcome association, and retained evidence.
• Heatmaps combine clicks and taps, scroll exposure, aggregate pointer activity, friction, outcomes, and comparisons. Pointer activity is not eye tracking.
• Journeys show branches, exits, time between steps, sources, campaigns, outcomes, and matching recording availability.
• Returning visitors shows who is new, active, came back, or went quiet, plus when and where return visits begin.
• Recordings let you watch consented sessions. Issue links seek to the relevant moment when an adaptive or continuously sampled recording exists.

Signals in plain English

Experience can surface these kinds of clues without exposing the internal detection rules:

• repeated clicks or taps on the same action;
• an activation that appears to have no response;
• something that looks actionable but is not;
• an interaction that responds slowly;
• a quick return after navigating away;
• repeated scrolling while searching for information;
• uncertain pointer movement or hesitation near an action;
• a control that moves while somebody tries to use it;
• form struggle or leaving after starting a form; and
• an action followed by a related browser or network error.

Treat a signal as a lead for investigation, not a verdict about a person’s intent. Confirm important findings with aggregate behavior, journeys, and available recordings.

Privacy and masking

Experience starts only after consent. Before consent, leadmaps makes no Experience configuration request, imports no Experience capture code, installs no listeners, and keeps no replay buffer. Revoking consent removes active listeners and clears active capture buffers and recording state for the page.

Experience summaries can include aggregate counts for interactions, exposure, scrolling, pointer density, broad gesture classes, and safe form progression. They never include typed values, keystrokes, page text, raw DOM, query strings, persistent device identifiers, or exact pointer paths. Password, payment, file, free-text, and editable-content fields are excluded.

Use blockSelectors to exclude an entire sensitive region. Exact pointer movement can appear only inside a consented, masked recording when replay is enabled.

Plans

Capability	Hobby	Pro	Business
Aggregate Experience capture and heatmap layers	Yes	Yes	Yes
Prioritized issues and richer journey or return analysis	No	Yes	Yes
Adaptive recording triggers and privacy-safe AI summaries	No	Yes	Yes

Anonymous privacy mode is excluded on every plan.

Performance behavior

The Experience code is lazy loaded, so it does not enlarge the core SDK entry. Movement and scrolling are sampled into compact aggregates, work is scheduled away from the browser’s immediate interaction path, and page summaries are strictly bounded. Lower-priority aggregate entries are dropped with visible overflow counters instead of allowing an unbounded payload.

The server processes issue grouping and impact analysis asynchronously so collector ingest remains lightweight.

Limitations

• A recording is not promised for every issue. Continuous replay is sampled, and adaptive recording depends on consent, plan, policy, quota, and browser availability. leadmaps retains no more than three adaptive clips per session; additional triggers are rejected or combined. The dashboard says “No recording available” when needed.
• Responsive layouts can differ. leadmaps separates incompatible layouts and falls back to a coordinate view when an element cannot be placed reliably.
• Touch devices do not produce ordinary mouse movement. Pointer patterns must be interpreted alongside clicks, scrolling, and outcomes.
• Cross-day returning behavior requires consent-mode persistent identity or an identified visitor.
• Outcome and revenue differences are associations between matched groups. They do not prove that an issue caused the result. When evidence is too thin, leadmaps shows “Not enough evidence”.