*THIS ARTICLE IS AUTOMATICALLY TRANSLATED, AND MIGHT INCLUDE TRANSLATION MISTAKES.
An important part of regional cycling policy is monitoring the effects of measures on cycling usage. Since before-and-after measurements with local bike counters are often lacking, there is limited insight into the impact of improvements in the cycling network. However, with data from Strava Metro, the use of new or modified routes can be analyzed since 2018. Although this data only applies to Strava users, it could serve as an indicator for the overall cycling trend on a route. It is, in any case, interesting to further explore the possibilities.
Research into Strava Metro data elsewhere in the world has shown positive results. It was found to reasonably predict changes in cycling usage on modified routes, with similarities to local counters. (https://www.sciencedirect.com/org/science/article/abs/pii/S0144164722000435).
An interesting aspect of Strava is the distinction between recreational and utilitarian bike rides. How do these groups respond to upgraded bike routes? Are road cyclists perhaps more sensitive to these changes?
You can find the full report at: https://www.track-landscapes.com/strava-fietsdata-provincie-utrecht. This article summarizes the key findings.
Mapping All Changes to the Cycling Network
The first question is: where in the province of Utrecht has cycling infrastructure been renewed between 2018-2023, and what kinds of investments have been made? We have made an inventory of the cycling routes that have been renewed during this period, based on various sources. This was quite a challenge, as there is no ready-made map or database that tracks this information. The result, based on changes in OSM (OpenStreetMap), local knowledge, data from the Cyclists' Union, and news reports, is a map with approximately 60 renewed cycling routes. Although it is likely not yet complete, it is probably the most comprehensive map of renewed cycling routes in the Province of Utrecht.
The routes are divided into several categories, of which especially the groups 'renewed to cycling street' and 'cycling path widened & renewed' are important for this study. These contain enough roads to make more general statements about the effects of the renewals.
All entirely new cycling paths (new connections) have also been mapped. However, these are less relevant for this study, as no before-measurement is possible.
Relative Change
Strava cycling data provides information from two specific target groups (long-distance commuters and road cyclists), but an important point to note is that the use of Strava fluctuates. Both the number of Strava users and the number of rides they register have significantly increased between 2018 and 2023, meaning that an increase in usage on a particular road does not always directly indicate more cyclists. Additionally, seasonal and weather influences play a role, with fewer cycling trips occurring in colder or wetter months.
Therefore, we should not look at the absolute increase in Strava usage on renewed cycling paths but rather whether the increase is greater than on other cycling paths in the province. In this way, we determine the relative development of cycling usage per route, in comparison to the total Strava kilometers per month in Utrecht.
Roads that were renewed between 2019 and 2022 have been included in the comparison, as sufficient before-and-after measurements are not possible for earlier or later periods. The exact completion date is sometimes difficult to determine, as roads were often closed in phases.
An Example
A good example of a significant improvement in cycling infrastructure is the Traayweg near Austerlitz, where the cycling path was widened from 1.5 meters to 3 meters over a length of 2.5 kilometers between 2018 and 2023. The old asphalt was replaced with smooth concrete.
In April 2019, Strava recorded 110 utilitarian rides on this path. By June 2023, this had increased to 280 rides (a factor of 2.54). However, during that same period, total Strava usage (measured in cycling kilometers) across the province increased by a factor of 1.78. The relative increase is therefore (2.54 / 1.78 =) 1.42, meaning that the Traayweg was used 42% more compared to other paths in the province.
The same principle applies to sports rides, where in June 2023, 5,665 Strava passages were recorded. The key is to observe in the graph whether the line after the completion of the works is higher than before (the y-axis starts at 0, and the scale is linear). If so, there is a relative increase.
The work on the Traayweg took place in July and November 2019 (also recognizable by the dips in the graph). The comparison of cycling usage before the renewal (2018, the complete year before the changes) and after (2020 - August 2023) shows a change of 163% for utilitarian rides and 180% for sports rides (an increase of 63% and 80%, respectively). The increase corresponds to the moment of completion of the improved route.
The opening of nearby mountain bike routes in 2020 may have contributed to the rise in sports usage, but an increase was already visible before this opening.
Although the Traayweg shows a clear change, this is just one example. The number of utilitarian passages is relatively small, meaning individual cyclists can have a large impact on the figures. Therefore, multiple cycling paths and streets were examined to draw broader conclusions about the effects of investments.
What were the results? Utilitarian and Sports
The displayed percentage compares Strava usage before and after the renewal. More than 100% indicates an increase in relative usage, while less than 100% indicates a decrease.
In the most common renewals (bicycle streets and widened/renewed bike paths), there is a relative increase in both utilitarian and sports cycling usage. Although the difference is small, sports cycling has generally increased slightly more than utilitarian cycling. On some routes, there is an opposite trend, where sports and utilitarian usage respond differently (one increases, the other decreases). However, in most cases, they show a similar development.
Bicycle streets
After completion, the use of bicycle streets averaged 112% of the usage before the renewal (an increase of 12%). This applies to both sports and utilitarian Strava usage. This increase varies between 101% and 128%, depending on the year of renewal. On an individual level, some bicycle streets scored lower, such as Bastionweg and Heemsteedseweg, which scored less than 90% in utilitarian usage. No bicycle street scored below 90% in sports usage.
Widened/renewed bike paths
Widened or renewed bike paths show an even greater increase than bicycle streets. Utilitarian usage rose to 121%, and sports usage to 141% of the level before the renewal. The Traayweg shows the greatest growth, mainly due to the significant widening from 1.5 to 3 meters. None of the renewed bike paths showed a decline in usage.
Further conclusions
Identifying additional specific spatial or infrastructural features (beyond ‘new bicycle street’ or ‘widened bike path’) that have the most positive impact on cycling usage is not simple and cannot easily be quantified, but it is still valuable to create an initial synthesis. Several features emerge from both individual examples and the overall percentage changes. These are discussed in more detail in the four examples in the main report.
In summary, we see that the largest increases in usage generally occur with significant improvements in the profile (width and surface) and the continuity of these improvements.
Bicycle streets show the greatest effect when the routing is also enhanced, with clear priority and a consistent profile maintained over a longer distance. Examples include the Panoven in IJsselstein and the Wittevrouwensingel in Utrecht.
With widened bike paths, a significant increase in width, for example from 1.5/2 to 3 meters, shows the greatest impact, especially for sports cyclists. This is clearly visible at the Traayweg and Fornhezelaan.
The Boerendijk in Woerden demonstrates that a new, wide cycling route, which replaces an unsafe shared road, also leads to a strong increase in usage.
In urban areas, the effect of bicycle streets is more pronounced than in rural areas. This is likely because cities offer more alternative routes, allowing shifts in usage to occur more frequently. Outside of cities, cyclists are more likely to rely on a fixed, direct route, with less opportunity for such shifts.
Change takes time
An interesting pattern is the speed at which usage changes after completion. On average, usage on most routes continues to gradually increase over the years.
For bicycle streets completed in 2020, we see that utilitarian usage in 2021 has not yet exceeded the level of 2018-2019 (only 89%). It is only in 2022 and 2023 that the previous usage levels are clearly surpassed. This pattern is also visible with other bicycle streets and widened bike paths: after the road is closed for construction, it takes some time for cyclists to rediscover the new route. Sports cyclists seem to be less affected by this, but even for them, usage gradually increases.
Also, with post-measurements using local bike counters, it is preferable to measure over a longer period: at least one year, preferably two.
What does this ‘effect’ mean in Strava cycling data analyses?
The meaning of an ‘effect’ or ‘increase’ is not always straightforward. It could be that the new bicycle street has made cycling more attractive, encouraging people to cycle more often instead of using the car or public transport. Another possibility is that cyclists have simply adjusted their route, with the new route replacing competing paths. In many cases, it seems to be a combination of both factors.
Ideally, we would compare changes in Strava cycling data with local bike counter data to better determine how representative these data are for general cycling usage. Unfortunately, such counts are often unavailable or unusable.
It is also possible that more people in an area started using Strava without a direct link to infrastructure changes. However, if this were the case, this trend should also be visible on other routes in that area. Our analyses show that this is only the case to a limited extent.
A lack of increase in usage does not necessarily mean the investment was pointless. Cyclists may experience more comfort without this directly leading to more usage, and we also consider this valuable.
Despite the limitations, the results support the belief that investments in cycling infrastructure have a positive impact on cycling usage and route choices. As shown by this study, Strava data offers considerable potential for measuring the effects of cycling investments. Further research, combined with long-term local counts, could provide even more insights.
Comments