This is neat, especially because I just started riding again this week after a few years away from it!
The analytics look good, and I wish I had them for running, too. I'm not sure how different those would be.
My one comment about the site now is that for someone like me who has only ever ridden for fun (or a commute), it could benefit a lot from having some basic information about the electronic gear. Reviews, or a couple suggested models, or even just a link to a good website.
For example, I didn't even know that bicycle power meters existed before just now. It took a few minutes of googling to discover that they cost around $1000.
I knew about bike computers, but I don't know much about them, except all my friends seem to have Garmin computers, and they all seem to hate them (reliability and user interface). So I'd be interested to read about what other options there are.
It doesn't need to be encyclopedic -- just a few hints to get me started. Apple used to show 3 columns: Good, Better, Best. I thought that was a good way to summarize what the differences were, and what they cost.
Thankyou! Running is something I don't know much about, but there isn't the same depth of analysis available as runners don't have power meters. I believe there are some tools out there for running that do more than just show runs on a map, but I couldn't point you to any offhand.
You're not the only person to suggest providing information about how to get set up with equipment would be a good idea, so I'm thinking about doing something about that soon.
Garmin dominate the market at the moment. The other main option is SRM head units, which are popular in the pro peloton, but cost around $800 and don't have a GPS. It's possible to use a smartphone if you have an ANT+ (the Bluetooth-like wireless protocol that most power meters use) dongle, but I haven't heard of any cyclists out there doing that (I'm sure there are some though). Specialized are coming out with an ANT+ bike computer next year, and there are probably some other options. To be honest, although there were some problems a couple of years ago with the Garmin Edge 500 (which I have), I quite happy with my device with the more recent firmware updates.
Some thoughts (coming from someone who has a triathlon in 3 days):
1. Great UI. The Garmin Connect team should take note.
2. Consider generating best fit curves for the graph data, it's hard to get an idea of how fast you're going over time, especially when you ride in the city and traffic and stop lights cause much of the spikey data.
3. I know the focus is power here, but it would be nice to get the Y-axis units to change when all that is displays is HR or Speed.
4. If you work in running and swimming I will throw money at you.
2. Adding optional data smoothing for the graph data is something I'd like to do soon.
3. The y-axis scales for power, speed and HR are independent, so it should work fine when there's no power data. Does it not work as well as I thought?
4. Unfortunately I'm not about to support other activities because I know very little about them, and at least while this is a one-man project, it pays to stay focussed. Hopefully someday...
#4 - have you looked at http://www.finisinc.com/swimsense/ (for swimming)? I chatted with them when I was involved in a swimming analytics startup and they are top notch!
Well done. Some feedback:
- People don't scroll. It wasn't obvious to me that there was more info below the fold.
- You need a strong call-to-action on the front page.
- Put this demo ride link above the fold, because showing the product is great: http://www.cyclinganalytics.com/ride/897403933578
- Sell benefits, not features and lead with that. How does our app make me a better cyclist?
- Consider showing a photo of a successful cyclist on the landing page.
The problem is probably more that there's too much white space between the image and the text, so it's not obvious you need to scroll on certain resolutions (iPad maybe as the text begins to appear on my laptop).
I don't always scroll, but because the images were cut off on the bottom of the page it was obvious to me that I needed to scroll. It may be a difference in configuration?
People who are in to cycling don't need to know benefits, they know that logging this data is important to them. They are looking for a way to do it.
My critism is different. Once I signed up I didn't really understand how to use it. Maybe it is my lack of experience with cycling apps?
Thankyou for both of your thoughts. I greatly appreciate this feedback.
I'm targeting a niche market with this, but I'm not quite sure about the level that I should be talking to them. On one hand, it's easy to assume they know a fair bit about the data that the power meter they spent a couple of thousand dollars on produces, so I don't need to explain what everything is. On the other hand, that's probably a dangerous assumption that often isn't true, so the more explanation I give, the better. I'll make the front page better in time, but I'm not entirely sure how to strike the right balance.
What part of using it after signing up are you having difficulty with? Have you uploaded any rides? Email me at david@cyclinganalytics.com if you like.
I'm at a stage where I am looking to add some measuring gear and upload rides to a system.
I recently upgraded from a bike that I maxed its potential (older, rebuilt road bike) to a new bike where I am nowhere near its potential and I'm looking to improve my training to get to the next level.
I registered, clicked around, but didn't know what to do. For those that have a garmin and experience entering data in an online system, it is probably easy. For me, I was left wondering what I needed to do to get started.
That is just my personal experience which may not be worth addressing until you get more scale. If you think your platform can take share from other data platforms fast, you don't need me yet (or customers like me).
If you think you need to grab advancing cyclists not yet measuring themselves, my comment is worth considering. If you are in need of advancing your platform through new users, my recommendation would be to add a link to 3 levels of equipment you recommend. Base (to get started, as low price as possible). 2nd Tier (something that give both a trip computer and power measuring device), 3rd (some advanced tier)
(This also may be a terrible suggestion since I don’t know enough about the equipment, like whether every bike needs different stuff etc.)
*Remember, the person paying $2,000 for a bike and $80/year for a service like this has disposable income, but maybe not disposable time so it may be worth making it really fast to get set up, including equipment.
Yes, what I have at the moment isn't very suited to people completely new to the field. The basic workflow is: plug your Garmin device into your computer, go to your user page (mine is http://www.cyclinganalytics.com/user/1000000), click the "Upload new rides" button, navigate to the Garmin device and find some ride files (probably stored in the "Activities" directory) and click okay.
Providing more information about how to get started with equipment and my site is a good idea and I'll look at doing something about that. Users having disposable income but not disposable time is a great point.
I'm not sure that you should explain how to do it, or even worry about my post. I may be a really limited number of potential users and you might have a much better way to spend your time.
If it were me, I'd focus on taking share to get a good number of users first, then worry about growing the market.
The existence of this might finally push me to purchasing a bicycle. Regarding the need for a bike computer and power meter would it be possible to include a few recommendations or suggestions, for people that have no idea about any of this? Might generate some Amazon affiliate revenue too.
One of the most critical information for "serious" cyclists is on rest/recovery intervals. None of the existing online tools have this, so people like me have to work on it offline and manually in a spreadsheet[1]. If someone could find a way to combine the effort of each training in the same chart, that would be awesome because it'd make recovery much more manageable.
Absolutely agree. Going hard all spring and summer is easy, but peaking for certain events (possibly multiple events) makes the scheduling of a training plan much more difficult.
The training load table[1] is designed to help with that. It doesn't yet let you put in numbers for future dates to help with planning, but I do plan on doing that.
Looks good. I uploaded a few rides. It would be nice if you could auto imports rides from Strava. Not sure if they'd be cool with that.
I bulk uploaded 23 rides, there was no clear indication when the upload was done and when I could navigate away from the page. I assumed it would continue processing rides in the background, but I guess not since I got an error and had to reupload my rides.
Thankyou for this feedback. I'll look into importing rides from Strava, so we'll see what happens with that.
It displays the little map and basic data for the rides when they're finished uploading, although now that you mention it, it's not obvious that you should be waiting for that rather than waiting for the progress bar to reach the end. And rides showing zero distance is something I haven't seen before. I'll look into that.
Privacy controls is a work in progress. At the moment, nothing is visible to anybody unless you give them a link to a ride, and then everything is visible to them, which certainly isn't ideal. Better privacy controls is near the top of my list of things to do.
I completely agree that a better way to upload rides is desirable. Are you thinking about the Strava upload interface?
The Strava upload interface is ok, but I find it slow. I think Garmin has a web plugin you can access from JavaScript, not really sure, I've never looked into it.
I've thought about writing a small app that runs in the background, and every time a garmin is plugged in, it grabs the latest files and uploads them to training peaks, strava, garmin connect etc.
It's non obvious how to find the Forve vs Cadence and Histogram graphs. If I didn't remember them from the splash screen I would have gone looking for them. Their text looks disabled.
I agree. The UI is very pretty, but some text is clickable and some isn't, and it isn't always apparent which is which.
Also, the intro message after signup is a fantastic idea, as is using question mark footnotes, but again, I didn't notice them until I read the message.
Interesting idea. How is Training load calculated? Some sort of algorithm that takes into account upcoming race schedule, age, category, past riding schedule?
And 'Pedal force vs. cadence', does this reveal where weak spots are in your stroke? Also does it show you your optimal cadence, based on speed and duration of ride. I think that would be great information to have.
Also, with strava I'd love to see when different sections are starting and ending so I know exactly when to gas it and stop, perhaps even how I'm progressing as I'm riding. This is pretty unrelated though to this app.
The training load is based on the exponentially weighted moving average of the normalised work of rides, where the normalised work is a number that shows how "hard" a ride is. A one hour time trial is worth 100, so if you do just that every day for a long time, you'll end up with a long-term stress and short-term stress of 100, and a stress balance of 0. There's a longer explanation over on the blog: http://www.cyclinganalytics.com/blog/2012/06/the-training-lo...
This isn't able to show any weak spots in your stroke because this data is only being recorded once every second. I think you need a Wattbike or similar to show exactly what your pedal stroke looks like. I'm not sure how to work out optimal cadence, but it sounds like something worth thinking about.
Are you talking about when you're using the Strava phone app? I haven't used it, but I'd definitely be in favour of it showing when you're starting segments and doing other smart stuff like that. I'm planning on doing a bit more Strava integration with my site too, starting with getting the Strava segments for a ride.
I've used a computrainer before to capture this data, weak spots in pedaling. Not sure how this technology works though.
Yes, Strava phone app. I don't have the fancy smancy garmins/power meters/etc. Does it show though on a Garmin when you enter a segment if you are running Strava?
It looks like Computrainers capture more data[1]. I'd love it if power meters on bikes also captured this data.
Nah, us Garmin users don't even have an app on our devices, we just upload our data to their website when we get home (just like it works with my site).
One feature I have wanted that would solve a frustration I have with Strava is the ability to quickly "prune" a ride the same way you edit a video. When riding I don't want to think about pausing the computer and other meters when taking breaks. Instead I want to have the option to record everything and then edit out the segments that I know were stops. The app could even autodetect part you may want to prune by recognizing when your GPS coordinates remain relatively static for extended periods or time.
>One feature I have wanted that would solve a frustration I have with Strava is the ability to quickly "prune" a ride [...] The app could even autodetect part you may want to prune by recognizing when your GPS coordinates remain relatively static for extended periods or time.
Strava does do this. If you look at a ride, note the difference between Elapsed Time and Time. Elapsed Time is the time you spent moving.
A few of my friends use Endomondo to track their cycling activity and get the statistics, share routes. It works nice and basically you just need a smartphone with GPS and a bike and you're ready to roll.
They're cycling hobbyist though but bike quite often.
There's definitely an opportunity to disrupt in this segment, although focusing on power is going to substantially limit your market appeal. For example, I'm a record holding cyclist and I've never trained power.
Yes, I'm aware focusing on power data limits my potential audience, but I don't yet regret doing it. 1) There are enough power meter using cyclists around to make this viable. 2) I'm not sure how to make this amazing for non-power meter using cyclists. 3) Strava is fantastic for more casual cyclists (not that that describes you). 4) I'm sure power meters will become cheaper and more popular over time.
You know how you can make it amazing for power meter and non-power meter users alike? Add a feature to help analyze tests. Help guide users to picking an appropriate time interval and course, and help interpret the results, and save those results to a profile.
And as power meters become cheaper and more popular, more people will be looking for ways to interpret that data. Some will turn to Hunter and Coggan's book or to a coach, but some will inevitably turn to sites such as yours for guidance. You already seem to have started down that path, but the more you can do to guide and educate inexperienced power meter users, the better you'll do.
Let me think about some ideas for #2, re: #1 and #4, I only know of maybe 5 cyclists in all my circles of cycling mates that use power (out of probably hundreds) and power meters always cost $2k every year the same way a great bike always costs $10k every year. :-)
oh, and regarding #3 - strava is good fun and bragging rights and mostly useless for any training purposes. If I understand your purpose correctly, you are really targeting a training segment, not a riding segment so Strava probably isn't a serious concern (until they decide to get into offering training features...)
Awesome site! I don't anything constructive to say except that this is exactly the kind of site I have been dreaming of. Unfortunately I lost my power meter on a ride about a year ago and haven't replaced it yet. If I do I will be sure to make good use of your site. I'm especially impressed with the graphs. I saw you mentioned you wrote the code yourself. Any plans to open source it?
Don't worry about the ugliness. If others find it useful you will find them submitting patches that take the ugliness away. And you've got more important things to worry about like adding new features and making the copy even better.
Hey David, looks good. Have you checked out our site, http://ridewithgps.com ? We've been around for a few years, are bootstrapped and profitable. Interesting to see opinions that there is still plenty of room for disruption in the market. It's always nice to get a new competitor, keeps everyone on their toes ;)
I have seen your site before, but I would hesitate to say we're competitors since you don't seem to offer much in the way of power analysis, and that's pretty much what I'm doing :-)
Some do, some don't. Most of the racers I know (people who actually train to race and take competition at least a little seriously) would know what this data is telling them or would know someone who could help them (a trainer or coach).
It's only free for the beta. If I was in the US, I'd be using Stripe and be charging for it already. However, I'm in Australia and I'm currently thinking about the best way to do this.
Javascript and a HTML5 canvas — I've written all the graphing code myself. I prefer to use stuff that other people have written, but I wanted flexibility, and not all graphing libraries handle tens of thousands of points well (and neither do SVGs).
Cool, well you got some nice looking graphs taking the path you did. I am working on an app that has graphing, and I am running into some of the problems you mentioned. I am using the highcharts library.
I haven't actually tried d3.js. I encountered significant speed issues when using Raphael (even when panning a graph), and I assumed that was more of an SVG problem than a Raphael problem. Maybe I shouldn't assume so much, but I didn't think d3.js would be suited to showing graphs with multiple data sets with potentially ten thousand points (or more, there's a data point for each second, and it's not uncommon to be riding for more than three hours) each.
Good idea. Garmin Connect doesn't yet have an API, so I don't think that is possible. It's probably possible to do it with Strava, so I'll look into that.
I'm a big fan of Strava, but it's not very useful when it comes to doing stuff with power data. Their premium accounts have some basic analytical tools, but I can't imagine any power meter using cyclists using that as their primary analytical app.
One of the first things I made my site do was upload rides to Strava, because I was tired of having to upload my rides to multiple apps.
This looks good. So many people are sick of WKO and Training Peaks.
Have you checked out Golden Cheetah? It's got most of what you have, plus it can control computrainers, etc. It's a great cross platform desktop app, but the UI needs some loving.
Thanks. I like Golden Cheetah, but I had a few itches I wanted to scratch... Amongst other things, having it as a webapp lets me use this on all of my computers.
Looks very nice (Golden Cheetah user with Garmin Edge 705 and PowerTap SL+ hub).
What about tools to help edit out unwanted sections of a recorded ride?
My type of riding, mainly long distance (Audax), means rides of 9 hours or more (up to 116 hours) with various length stops for food and/or naps (for the multi-day rides). It was a lot of faff with Golden Cheetah (or even editing the .tcx files directly) to trim out sections where I'd stopped and accidentally left the Garmin recording.
Some nice way of selecting a chunk of a ride (whilst either looking at a zoomed in map or selecting a section in the speed/power/cadence/HR graph) and then moving it to a new 'lap' (where it could be kept or deleted) would be very useful.
Cropping rides is something I'll be adding soon, but I hadn't thought about removing sections. Actually, at the moment, I'm essentially treating rides as arrays of records at one second intervals, so having gaps in the middle doesn't work. To be honest, I hadn't considered the use case where you have a single ride file that legitimately spans multiple days. I'll think about this.
Thanks, that probably explains the behaviour I'd emailed you about. Niche use case though, there aren't that many of us long distance loonies out there (The Great Southern Randonee 1200km late October starts quite close to you: http://www.gsr1200.com.au/gsr1200/Home.html ).
Let me know if you need some example large data files to test scalability. I've got one .tcx file (with heart rate, cadence and power data) that's 41MB (would have been 53MB if I hadn't had to reset the GPS part way through).
The analytics look good, and I wish I had them for running, too. I'm not sure how different those would be.
My one comment about the site now is that for someone like me who has only ever ridden for fun (or a commute), it could benefit a lot from having some basic information about the electronic gear. Reviews, or a couple suggested models, or even just a link to a good website.
For example, I didn't even know that bicycle power meters existed before just now. It took a few minutes of googling to discover that they cost around $1000.
I knew about bike computers, but I don't know much about them, except all my friends seem to have Garmin computers, and they all seem to hate them (reliability and user interface). So I'd be interested to read about what other options there are.
It doesn't need to be encyclopedic -- just a few hints to get me started. Apple used to show 3 columns: Good, Better, Best. I thought that was a good way to summarize what the differences were, and what they cost.