With the revamp, the Trailmap app moved to what is known as a freemium model, which means that the basic features of the app remain free of charge ("free"). The more advanced features that an active user wants ("premium") are available by buying the Plus package, which is a fixed-term subscription.

The move to a freemium model has drawn comments both for and against. Some users understand the need for the change and reckon that, although paying always stings a bit, the reliable continuation and development of the app you get in return are worth more. Others have raised various questions about charging for it. This piece goes through those questions.

Background and history

To understand the thinking behind the new direction, it helps to go through a bit of history first. In 2013 Antti Nissinen kicked off OSM trail mapping in Tampere and also built the first version of the "MTB Kartta" service. In 2016 Tapio Kaasalainen produced a version of the "MTB Kartta" style that fit Garmin devices, and later also reworked the general "MTB Kartta" style.

In 2018 yours truly built the "Bikemap" app, backend servers included. The motivation was to try out new software technologies hands-on and to get for my own use the features that Antti's stripped-down website did not have. Bikemap gradually accumulated new features, I opened it up to others too and the number of users started to grow accordingly. As use expanded to sports beyond mountain biking, I changed the app's name to "Trailmap". Along the way my own list of motivations picked up an aim wider than my own needs: to support both the mapping and, through the map, the people who get out into nature, on a bike or in running shoes, and find new experiences on the trails — this is still Trailmap's mission.

The year 2023. The community has mapped & (mtb-)classified over 23000 km of trails into OSM (in Finland)! It has been great to watch OSM trail mapping take off and grow, especially after the "big leap" of 2020. More and more of the mtb riders, gravel riders, trail runners, dog walkers and so on out on the trails have got used to making use of this "treasure trove", and my guess is that most of them do it using Trailmap. Even though there are still (bafflingly) plenty of people who believe the NLS topographic map tells you where the trails are… If Trailmap has a bug, is sluggish or is missing a feature, word is sent to FB or straight to me by email pretty smartly; for many, Trailmap is an app just like Strava, Komoot, Garmin or whatever, with the same expectations about reliability, quality, features and so on.

On this side of the table, though, at this point in the journey the situation started to give me pause: it has of course been great to see a hobby project take off and benefit lots of people — but on the other hand it was never supposed to become a widely used service like this, given that its technical implementation was a cheerful hodgepodge of technology experiments produced in the name of a hobby. So maintaining it had already become work, and it kept drifting further from the self-serving hobby the project started out as. A service that is free for users is anything but free to run, and when you try to keep costs from growing, the amount of maintenance just keeps growing while reliability and performance stutter. Running a service like this is interesting as a hobby in one particular way: you rarely get to choose when the hobby happens. If there is a problem with the service on a Sunday morning when you were about to head out for your own ride, the problem is right there in that moment, and it does not wait for you to get your ride in first. And to be clear: there is no "team" here sharing the workload — this is a classic one-man-at-the-lathe project.

How to go forward? One option would be to say "all good things come to an end", scale back or wind up this hobby project and put the freed-up time, energy & money into the other chores competing for a limited amount of time. The other option would be to make use of the lessons picked up along the way, of the OSM trail mapping that has got up an excellent head of steam, and of the user base that has grown, and rebuild Trailmap on a more sustainable footing. At this point it is probably clear which path is currently under way.

Trailmap's big revamp 2023 - 2024

What does this revamp mean in practice? The "hodgepodge" — i.e. many central parts of Trailmap's implementation — had to be rebuilt so that developing and maintaining them further would be worthwhile, and server capacity had to be added and servers duplicated to ensure they keep working. These answered users' cries for reliability, response times and new features. In practice this required a long-running, significant investment of time and money — including the duller sort of graft that tends to be called "work" rather than a hobby.

As the goals of this new project and the work and money it demanded started to become clear, the thinking had to shift to how to both make this possible and make it worthwhile for myself in the long run. The solution was to acknowledge that developing and maintaining Trailmap was no longer purely a hobby but had both a "work" and a "cost" component that would need compensating. The answer that emerged was moving to a "freemium" model, where occasional use — looking at the map, say — would still be free, while regular use carries a fee that corresponds to the benefit the user gets from the app.

A central feature of the freemium model is that it still lets Trailmap remain a way for a wide base of enthusiasts to make use of the OSM trail map. With Trailmap anyone can also share routes, trusting that others can open the shared route without fees or requirements to install an app or register as a user. If, on the other hand, someone's occasional use of Trailmap becomes regular and the app becomes a central part of their sport, then to my mind it is reasonable for the user to take part in the cost and the work of developing the app and running the service by paying a monthly fee of around three euros.

Trailmap will still be developed fundamentally "for the love of the game". The freemium model is an enabler: I can put considerably more money and time into developing the service and its quality than would otherwise be possible, and I can make decisions about the app's future for the long term.

Questions and discussion

When the freemium model was first discussed among the OSM mappers, a few questions and views came up that are worth documenting here as well.

Most of the mappers shared the concern about securing Trailmap's continuity and saw the freemium model as workable, hoping that the basic feature, "viewing the trail map", would remain free. Some mappers, on the other hand, found charging for it — or the "commercialism", as some described it — difficult. The questions and views raised by those mappers can be simplified into three themes:

• Is it fair for Trailmap to "cash in" on the work mappers did for Trailmap for free?
• Mappers do their work for free, why does Trailmap's developer not do the same?
• The basic features, such as map tiles, should be free because 1) they are based directly on the work mappers did and 2) apps like Strava offer plenty of basic features for free in exchange for the user's data, and then you can pay for "premium" — but the premium has to be clearly something other than the map itself.

My comments below approach those themes from different angles, while also straightening out a few of the more general misconceptions that came up in the discussion:

1. OSM (trail) mappers have mapped into the OpenStreetMap community map (OSM), not for Trailmap. Trailmap does use OSM data, just as many other apps do (Komoot, for instance, uses exactly the same data and the same mtb:scale tags to classify trails for mountain biking). So Trailmap has no exclusive right to the data, nor any particularly good terms: OSM's ODbL licence strictly regulates what you may and may not do with OSM data. That is fine, of course, but it probably clarifies further the question of why the result of the mapping really is not "Trailmap data".

2. Mappers make their contributions for free, so why does Trailmap's maker not also do all of his work for free. This equates the contributions of the mapping community and its individuals with the development and maintenance of Trailmap. There are similarities, as I noted above: both have the typical features of a "hobby", i.e. interesting, self-directed activity that is rewarding directly through the doing — the mappers map and I develop new features into Trailmap — such that both the doer and the community benefit from the results in either activity. This would apply to Trailmap 1:1 with mapping if Trailmap's development stopped at what is required for my own use and others could then use the same thing on the same terms — which is exactly what Trailmap was at the very start ("Bikemap").

   Today's Trailmap is a long way from "making it for yourself": once a feature is good enough for my own use, there is still several times that amount of work ahead to make the app fit for a large user base: polishing the interface for different devices and for beginner users, testing, chasing bugs, building reliability, writing help pages, providing user support. The service also requires considerable server capacity so it runs for a large user base — this costs both money and time. You can map whenever you want, as much as you want, and take breaks whenever you feel like it — running a continuously operating service like Trailmap is a responsibility that does not take holidays. The service is expected to run every day of the year, and any problems have to be responded to quickly.

   Note: none of this belittles mapping in any way; it only makes the differences between the activities concrete. A key factor in the success of OSM's community mapping is exactly all the positive features that have been built into it, and with them an enormous community of mappers has been built. Correspondingly, one of OSM's biggest challenges is advancing the things outside mapping: guidance, quality assurance, tools, software libraries and so on — taking those on is not felt to be as inspiring as mapping and they are chronically without takers.

3. "Trailmap would not exist without the mappers' work..." was an argument used a couple of times in the discussions. On its own this is factually correct, but using it as the main basis for arguing that Trailmap should be free for everyone misses the mark. Trailmap is an app / service that makes use of OSM data, just like Komoot, Strava, MapMyRide, Suunto, Garmin and so on and so forth. All of them have benefited from the mapping just the same — that is the central idea of open data (which OSM data is). Trailmap's difference is that it has also actively supported the mapping work, not just made use of its fruits.

4. Comparisons to Strava's free features, or "the map at least should be free in all its forms", seem to be based a bit more on gut feeling than on any analytical thinking about how value is created and exchanged. Strava, for instance, like many similar services, gets direct value from every free user who shares their data with Strava — that user data becomes data owned solely by Strava, which Strava can monetise in many ways. Trailmap has no equivalent mechanism by which it benefits from users. The "the map should be free..." comment is also worth comparing to OSM's own practice: the openstreetmap.org web service is free, but there are clear rules concerning OSM's tile servers ("OpenStreetMap data is free for everyone to use. Our tile servers are not.", as their fair usage policy puts it) — and this despite the data being owned by OSM and licensed by it to others.

5. The comments about "commercialism" have carried an assumption that developing an app like this would be significantly profitable business. Trailmap's target market is, however, quite limited in size and the work required is considerable — a simple calculation tells you these realities. So the project cannot be powered by potential business profits, but by a passion for solving the problem itself, "for the love of the game". The freemium model is the enabler for the venture: it makes sure the developer's wallet does not empty out mid-project, and that a staggering amount of work does not get sacrificed on something users hardly value.

6. The MTB trail map (i.e. Trailmap's base map) is within anyone's reach: OSM data is open, as noted above, and the map style (Carto CSS / Mapnik) and the map server source code needed for the MTB trail map are available on Github: https://github.com/MaRaSu/openstreetmap-tile-server.