Measuring whether your map is actually working

Once a map ships, the temptation is to stop measuring it. The interesting part — drawing, naming, sharing — is over, and the analytics dashboards all look like noise. So most map owners don’t look at their numbers at all, and find out the map is in trouble only when somebody complains.

This is solvable. There are exactly four numbers worth tracking, two numbers that look important but aren’t, and one twenty-minute review you can do once a month to catch most problems before they compound. That’s the whole guide.

The four numbers that matter

1. Returning visitors

The healthiest signal an interactive map can produce is people coming back. A map that nobody returns to is either a one-off curiosity or a map that solved its problem in one visit. Either is fine for a static reference; for a living map, returning visitors are the whole game.

Track the percentage of visitors who come back within 30 days. Healthy community maps land around 25–40%. Below 15% and the map is being consumed once and abandoned — usually a sign that it’s static when it should be evolving.

2. Time on map

Time-on-page is a misleading metric for most websites and a useful one for maps. A map that’s working keeps people exploring; a map that isn’t gets a quick scan and a close. Aim for a median session of 60–120 seconds; below 30 seconds means viewers aren’t engaging, above 5 minutes can mean the map is so dense people are lost in it (which is a different problem).

3. Region click-through rate

What percentage of viewers click on at least one region during their visit? This is the “is anyone interacting with the actual content” metric. Healthy maps land at 50–70%. Below 30% means either viewers don’t realise the regions are clickable (a UX problem) or the regions don’t have anything interesting behind them (a content problem).

4. Submission rate (if you accept submissions)

For maps with community submission enabled, the rate of incoming submissions tells you whether the contributor base is growing or dying. Track submissions per week, not in total — the trend matters more than the absolute number. A flat or declining trend over six weeks is a leading indicator that contributors are losing interest; a rising trend is the opposite.

The two numbers that look important but aren’t

Total page views

Total page views go up over time as a function of basically nothing — word of mouth, search engines indexing the page, somebody pinning the link in a busy channel. They feel like growth but tell you almost nothing about whether the map is working. A map with 50,000 views and 5% return rate is in worse shape than a map with 5,000 views and a 35% return rate. Track returning visitors instead.

Unique visitors

Same problem. A map can have huge unique-visitor numbers because it got linked from somewhere big, then deteriorate quietly while the headline number stays high. The signal is in repeat behaviour, not in the breadth of one-time exposure.

The monthly review (twenty minutes, end of month)

Block twenty minutes once a month. Open the analytics dashboard. Walk through these five questions in order:

1. Is the return rate trending up, flat, or down vs. last month? Down is the alarm signal. Look for what changed — new visual category, removed feature, drop in submission cadence.
2. Is the click-through rate trending the same way?If both return rate and CTR are dropping, the map is becoming less engaging — usually a content freshness problem.
3. Which regions get clicked the most? The top five regions almost always carry a disproportionate share of the engagement. Make sure their content is current and accurate; the return on polish is highest there.
4. Which regions get clicked the least? Bottom-five regions either need better naming, better placement, or to be merged into a neighbour. They’re cheap to fix and the cleanup compounds.
5. If you accept submissions, how many came in this month? Compared to the rolling three-month average. If it’s falling, the form is too vague, the SLA has slipped, or you’ve been silently rejecting too many — investigate.

That’s the whole review. Twenty minutes, five questions, no spreadsheet acrobatics. The discipline isn’t in the analysis — it’s in actually doing it every month instead of skipping it when nothing seems urgent.

What good looks like

A healthy interactive map, six months in, looks roughly like this: 25–40% return rate, 60–120 second median session, 50–70% CTR, a predictable cadence of 5–20 submissions per month with a 15–40% rejection rate. Numbers in those ranges mean the map is doing its job. Big deviations on any single one of them are worth investigating; deviations on two or more are a structural problem and worth a rethink.

The metric that doesn’t show up on any dashboard, but matters most: do you and your contributors enjoy maintaining the map?A map that nobody enjoys maintaining is a map that’s about to die, regardless of what the numbers say. If the numbers are fine but the energy isn’t, fix the energy first; the numbers follow.