I Own My Data

I've never like giving up control of my data and how it's presented. You can see that philosophy in how I run this blog, and how I design most of my projects. From the beginning, I knew I wanted a place where I could share my thoughts, ideas, and projects, and I knew it wasn't going to be something like Medium or Substack. It's important to me that my corner of the internet can live on with very little control from any platform, company, or person. Even if this iteration of my site is eventualy upgraded, replaced, migrated, or otherwise changed, it's always going to be in the same place. I host it on Vercel and built this version with Next.js, but platforms and frameworks can come and go. My content is all written in Markdown and it's pretty portable if I need to rebuild my site. In fact, I've done it a few times before.

I have complete control over most of the content and media here as well. Images are served from Cloudflare R2 for now, but I can always move to a different storage platform. Videos are served from my own infrastructure as well. I've avoided embedding YouTube videos or other external content, since I don't control what happens to my videos, and whatever provider I'm using could deactivate my account or change how it works, breaking tons of historical content.

I have had one gap in my strategy though, and it's a pretty big one in my opinion.

I Don't Own My Fitness Data... But I Should

For a long time now, I've been excited about connected fitness. I've also recently redeveloped a love for cycling and gotten pretty into road biking.

In order to visualize, share, and otherwise analyze the results of my workouts, rides, hikes, or anything else that I do, I have come to rely heavily on third parties. Strava is one of the main platforms that I use, both to share my rides with fellow enthusiasts, or to keep track of my training over time, and just look back on it for fun! This works fine, but given my thoughts on data ownership, it always bothered me that I didn't have more control over how I presented that data. Originally, I had started building a side project that used the Strava API to display my activities and embed them here on my blog. Then, I got an email a couple days ago (emphasis mine):

Hello Keegan

We’re reaching out to inform you of a change that affects how third-party apps connected to Strava may display certain information. This update, effective on November 11th, is part of our commitment to privacy and transparency across all connected apps and devices.

Third-party apps will now only be allowed to display Strava activity data related to a specific user to that user. Partners will be required to update their app within 30 days of November 11th to align with this new standard.

They expanded a bit on the changes in a press release, but this severely limits what developers are allowed to do with Strava data, including projects like the one I was working on. This change left me feeling bad, and got me thinking about fitness data ownership. I'm the one who dragged my ass up a mountain on my bike, and I'd like to do what I please with the data I collected and uploaded to Strava.

Taking Back Ownership

Now, I'm not necessarily saying that Strava is holding my data hostage or not letting me analyze it, but I don't like the level of control they do have, and this change was the push I needed to finally work on applying my same philosophy of data ownership to my fitness data. My first focus has been on presenting ride and hike data here on my site. I could use Strava's embed code, but that's still putting all the control in Strava's hands. If a month from now, or 5 years from now, they decide to disable embeds, all my old content will be broken, and these recent changes (especially with the short notice period) don't give me confidence in that regard. So, I built a library and component that can take my raw workout data and embed it here on the site!

The component works by downloading and parsing a GPX file, a format used for storing workout route data, including GPS location, elevation, heart rate, cycling power, and more. I can host these files myself, and the component doesn't rely on Strava, or anyone else, to display or calulate the data. The only service I rely on is Mapbox, but I can always swap them out for a different mapping service if needed.

I was even able to add some features I've been longing for, like showing the virtual map of Watopia for Zwift indoor rides, rather than overlaying the GPS track in the ocean like Strava does. This project will continue to evolve over time as I expand the features, support more workout types, and improve automation for getting things uploaded and stored on my own infrastructure. The end goal is to be able to have a feed of activities here much like Strava, but that I own and control. With video and photo embeds, I'm confident I can deliver an even richer and more flexible experience than many of the big platforms offer. At least I can build it to precisely support my needs and desires.

Zwift, the virtual training platform I use when I can't ride outside, lets me record videos and photos during the ride. Here's an example:

I have big plans for this new integration with my raw fitness data, including map overlays, data graphs, and rich media like photos and videos. Having full control has always been my goal, but recent developments just encouraged me to finally do the work, or at least get started.

Final Thoughts

Strava has made a strange decision in my opinion with these new API restrictions and the tight deadline to comply. Maybe we'll see them launch a new AI-based training platform soon, which I think would explain some of the changes. The reason doesn't matter though, because Strava is not unique. These kind of changes can happen, and will happen, when control of your data is in the hands of a company seeking to maximize profit. If anything, I'm reassured that I'm on the right path with the level of ownership I take in my content. It's more work, but worth it in the long run.

How do you feel about data ownership? I'd love to hear about the lengths you've gone to own your data, be it fitness data, blog posts, or anything else! Come find me on Bluesky and let's connect!