How to use a heatmap to find roads locals actually ride

Every town has roads that cyclists use and roads they do not, and the difference is rarely visible on a map. A heatmap makes it visible. It is a layer drawn from the GPS traces of everyone who has ridden there in the last year, and it answers, better than any guidebook, the only question that matters when you arrive somewhere new: where do the people who live here actually ride?

This is how the heatmap is built, what it is telling you, what it is not, and a method for turning it into a route in fifteen minutes.

What a heatmap is

The most widely used one is Strava's Global Heatmap. Strava's own description: it is built from aggregated, de-identified public activities over the past year, it is updated monthly, and no heat is shown in an area until multiple users have uploaded activities through it. It is free to everyone on the web version. You can filter it to a single sport, so a "Ride" layer shows only cycling. Subscribers also get a weekly heatmap, which uses only the last seven days, and a night heatmap, which isolates activity between sunset and sunrise, plus a personal heatmap of their own rides. Activities set to private, or from users who have opted out, are not included.

Two properties follow from that. First, the heatmap is a count, not a rating. A bright road is a road many people rode, not a road many people enjoyed. Second, it is a sample of people who record their rides and share them, which is a particular slice of cyclists: more sporting, more male, more urban and more affluent than the population that rides bikes. Pritchard's systematic review of revealed-preference methods in cycling research makes this point about crowdsourced data generally: the sample is large and cheap, and it is not representative, and the two facts have to be held together.

What the research says riders choose

Transport researchers have been reading GPS traces from cyclists for fifteen years to answer the same question you are asking. The findings are consistent enough to use as a lens.

Broach, Dill and Gliebe tracked 164 cyclists in Portland over several days and built a route-choice model from where they actually went. Riders were sensitive to distance, to the number of turns, to slope, to traffic volume on the roads they used, and to how intersections were controlled. They would go out of their way, meaningfully, to avoid busy roads and to use separated infrastructure. Skov-Petersen and colleagues found the same pattern in 1,267 GPS trips from 183 Copenhagen cyclists, in a city with far better infrastructure. Zacharias and Zhang, in a Chinese city with GPS trackers and intercept surveys, found tree shade and traffic speed among the strongest predictors of which streets riders picked.

Read together: a heatmap is not just where riders go. It is a map of what riders avoid. The bright line on the minor road parallel to the dark main road is not an accident; it is a hundred people making the same judgement about traffic that you would make if you knew the area. That is the information you are buying.

Reading the map: five patterns

• The parallel road. Wherever a main road is dark and a minor road beside it is bright, use the minor road, and trust that judgement even when the main road looks quicker.
• The bright climb. A single road glowing up a hill with nothing else around it is the local climb. It is bright because people repeat it, which usually means the surface and the gradient are good and the traffic is not.
• The dark gap. A road that is bright at both ends and dark in the middle has something in the middle: a gravel section, a gate, a motorway junction, a tunnel. Look at it in satellite view before you route through it.
• The dark valley road. If the obvious road up a valley is dark and a harder road over the ridge is bright, the valley road is probably unpleasant to ride. Locals do not climb for fun on a Tuesday; they climb to avoid something.
• The loop. Where several bright roads join into a circuit, that is the club loop. It will be the right distance for an evening, it will have coffee on it, and it will be ridden in one direction more than the other. Loops beat out-and-backs for a reason.

What the heatmap will not tell you

It will not tell you why a road is bright. A road can be bright because it is a lovely quiet lane, or because it is the only road out of town, or because a sportive with three thousand riders went down it last May. Filtering to the ride layer removes runners and hikers but not events. The weekly layer, if you have it, shows whether a road is used routinely or was used once.

It will not tell you about surface. The gravel riders and the road riders are in the same layer. A bright track through a forest may be perfect on 40 mm tyres and impossible on 25s. Cross-check with satellite imagery or with the surface information in a routing tool that carries it.

It will not tell you about time of day. A road that is fine at six on a Sunday morning, which is when the bright traces were recorded, can be a lorry route at five on a weekday. The night layer is a partial answer; local knowledge is the full one.

And it will not tell you about roads nobody records on. In much of rural Europe the people who ride most are older, ride without a phone, and are invisible to it. The dark road can still be the good road. The heatmap tells you where the confident local sporting riders go, which is usually a good enough proxy for a visitor, and occasionally misses the best lane in the district.

A fifteen-minute method

1. Open the heatmap, ride layer only, and zoom out until you can see the whole area you could reach in the time you have. Look for the shape of the bright network before you look at any single road.
2. Find the club loop. The brightest circuit near your base. Note the direction: if you can see where it starts and ends at a café or a car park, you can guess it.
3. Find the local climb, the bright dead-end or over-the-top road with the most traces. That is your hard effort for the week.
4. Draw the route on the bright roads only. Every time you are tempted to use a dark road to shorten it, ask why it is dark. Sometimes the answer is that nobody needs it. Often it is that nobody wants it.
5. Check the dark gaps in satellite view. Then check the profile.
6. Save it. In Moveee the Route Builder shows the profile and lets the Route Engine match a saved route to what a training session needs; a saved route can also be ridden in Moveee Indoor from a smart trainer if the weather turns before you get to it.

The method works at home too. Most riders have a heatmap of their own that is a handful of routes worn into a groove, and the Global Heatmap next to it shows what everyone else in the same town has found that you have not. In our experience the road you have never ridden that is bright on the map is worth an afternoon almost every time.

Where it fits

The heatmap is the first tool, not the only one. Finding good roads somewhere new covers the rest: what to ask a bike shop, how to read a topographic map, when to trust a routing algorithm and when not to. And road choice and safety is the longer argument for why a slightly longer route on quiet roads is almost always the better one, which is the argument a heatmap makes for you, in colour, without saying a word.