Read your tour analytics and fix the drop-off

Analytics only pay off when they change what you do next. Mango360 records what each visitor opened, how long they stayed in each scene, which hotspots they clicked and where they left — and the drop-off point is usually the most valuable number in the whole report.

This tutorial walks through the report, the order in which to read it, and the two changes that resolve most of the problems it exposes.

The five numbers that matter

• Opens: unique tour opens in the selected period, with visitor IP addresses hashed
• Scene views: how many times each scene was displayed, which shows the real structure of the tour
• Dwell time: average seconds spent in a scene; a low value usually means the scene is uninteresting or hard to read
• Hotspot clicks: which interactive elements were used, separated by type
• Drop-off: the scene where the last visitor of a session left, expressed as a share of entrances

Find the scene that loses visitors

1. Open Analytics and set the date range to a full week, because weekday and weekend traffic behave differently
2. Sort the scenes by drop-off share and find the first scene that loses more than a third of its entrants
3. Open that scene in the editor and look for the obvious causes: a low camera angle, a blank wall, no hotspot, a broken label
4. Check the dwell time of the scene before it: if it is high and the next one is low, the problem is the link, not the destination
5. Publish a fixed version, keep the previous one as a comparison point and check the same numbers the following week

Count your own events with postMessage

When an embedded viewer has to report into your own analytics, the postMessage API exposes the same events to the hosting page. The events carry the scene id and the dwell time in milliseconds, so your numbers can be compared with the ones inside Mango360 directly.

window.addEventListener("message", (event) => {
  if (event.origin !== "https://studio.mango360.app") return;
  const { type, payload } = event.data;
  if (type === "mango360:scene-view") {
    analytics.track("Tour Scene", payload.sceneId, {
      dwellMs: payload.dwellMs,
      sceneIndex: payload.index
    });
  }
});

Export, compare, repeat

Every report can be exported to CSV, which is the fastest way to compare two versions of a tour or to feed a monthly report. Keep one file per version and one per month: the columns are stable, so a spreadsheet comparing them stays readable for years.

• Export the same date range before and after a change, not two different weeks
• Compare drop-off share rather than absolute opens, because traffic volume moves on its own
• Put the version id in the file name, so the comparison survives a rollback
• In a multi-tenant organisation, compare across properties only after checking that the embed placement is identical

A useful rule of thumb: fix the first scene that loses a third of its visitors, leave everything else alone for a week, then look again. Tours that are edited in small steps, one scene at a time, improve far faster than tours that are rebuilt completely every month.