Creating a free, Strava-like, app

The $60 a year would cover a fraction of an hour of dev work on this project. As a dev the cost versus benefit of building something is off the charts.

As others have noted you have significant hosting costs. You pretty much need funding as Ads will not support this model short term. If you have a garmin you already have the option of garmin connect which does almost all of the same things as strava already and is free.

Open source ... if you can find a core group of people to donate hundreds of hours of their time for free you might have something in 2-3 years which still needs a hosting solution.

I develop an app that uses Strava data. It's free for now, but it is not free to run. It currently costs me about $25/month to keep it going. Over the last couple of years my usership hovers at around 5000 people, mostly in Japan and the U.K.

I have gone back and forth about having it be open source, but ultimately I want it to be open source. The problem is that this is my first real programming project and I wrote much of it not really knowing what I was doing. I learned everything from online articles and tutorials and Stack Overflow. The codebase needs a total overhaul to be something other people will want to work with, and that's what I've been working on lately.

Heatflask has relatively small user-ship. I have a few heavy users but it's not the kind of app most people will use every day. Data storage is cheap at first and the costs go up fast. That's true of any cloud services: they make usage cheap or even free for the hobbyist, but as soon as you go up a tier in performance or capacity, it can cost a lot. For example, Strava limits API calls. I get 5000 free API calls for every 15 minutes. So, if for example if three users want to look at all of their 1000+ activities (not unusual), that knocks out my app from importing data for 15 minutes. Every time my app requests info for a user or a data track, segment info for one user, that is one API call. They limit or charge money for that because it costs them money to maintain servers that are getting hit by everyone's API calls. So I have to at least cache peoples' data, so that if one user's data is being requested a lot I can pull it from my own cache storage. I'm using Atlas MongoDB for that.

Since I use maps, I have to consider the map tile service. Heatflask uses raster tiles for the maps. Tiles are 256x256 or 512x512 images that make up the map background. There are some free services but they also charge if you use a lot. Mapbox for example limits their free usage to 1 million tiles per month and they start charging after that. So I have to cache tiles as well to cut costs, but rather than cache tiles on a server, I cache them on the user's device.

As for storage of GPS tracks, you need to consider how you are going to be searching through them. You can store individual points so it's fast to look up any tracks by location but then it takes way more storage than a compressed track stored as one block.

Of course all of this stuff becomes way more complicated when you deal with data on Strava's scale, which is thousands of data transactions per second.

A Hungry, Violent Mob wrote:

This thread exemplifies the LRC Asperger's mindset fairly well. You'd rather spend potentially hundreds of thousands of dollars to develop a Strava clone than pay $7/mo for Strava. Got it.

This has done nothing except reinforce my love for Strava.

wait, what?

I'd really rather not see terms like that used to insult people. Just imagine how individuals with ASD diagnosis feel when they're the punchline of a joke that isn't even about them. I don't really care if this isn't a normal opinion on this forum - using ASD terms to insult people is not cool.

SJW warriors unite and burn the offender at the stake.

Trail Zealot wrote:

I'd really rather not see terms like that used to insult people. Just imagine how individuals with ASD diagnosis feel when they're the punchline of a joke that isn't even about them. I don't really care if this isn't a normal opinion on this forum - using ASD terms to insult people is not cool.

How was that the punchline of the joke? Did I LOL? Did I say knock, knock? Did I say what do you call a...?

LRC is full of socially abnormal people, probably many of which are on the spectrum. This is an example of the kind of thread that happens when you have no worldly experience and can't function within a society. Someone even said "it couldn't be too hard to create a Strava-like app." Really? How many developers you know that are working for free? Pay the $7 for Strava or not, but this is an preposterous idea.

Sorry that you find talking about ASD insulting.

Just chiming in with modern-day best practices for respecting people. You don't have to change anything you're doing, but I woulda felt sh!tty not speaking up.

I don't know anything about hosting costs, but how much storage would something like this take up and how much would it cost to host?

Based on my app's storage method I'm seeing an average of 5K (bytes) per activity.

I'm storing streams of 3D data points. Latitude, longitude, and timestamp. Lat and Lng are 32-bit floats stored in Google's polyline format and timestamp is a 16 bit integer RLE stream of time-differences. Any additional dimensions like altitude, heart rate , cadence, power, and whatever other sensors you might be monitoring would be stored like that too.

I don't think Strava can store them more compactly than that.

It's too hard to answer without doing some analysis. I mentioned Google Clould Firestore, which is what they call a NoSql database. Last I knew (summer 2019), they didn't actually charge specifically for storage, it was charge by access (reads, writes, updates). So most of your data "storage" wouldn't have a specific cost, but the reads and writes add up fast. Yea, you do have to be concerned with various privacy laws and data safety, but that is not an issue with the big service providers like Google, MS Azure, AWS.

Btw, if anyone is interested in experimenting and doing an app in Dart/Flutter, feel free to ping me at ians dot coding at gmail as I've been playing and would be interesting to see what could be possible collaborating.

Writing the code would be time-consuming but not overly difficult. The real challenge, as others have mentioned, is the infrastructure. You need to pay for bandwidth, processing, and storage, and none of that is cheap.

You also need to be able to scale. Scaling cost money.

and, lastly, you have to offer something Superior to everything else that's out there. I don't like the new charges anymore than anyone else but for $5 a month it is cheap enough that switching to an entirely new ecosystem would probably not be on everyone's to-do list.

hdbsydb wrote:

Writing the code would be time-consuming but not overly difficult. The real challenge, as others have mentioned, is the infrastructure. You need to pay for bandwidth, processing, and storage, and none of that is cheap.

You also need to be able to scale. Scaling cost money.

and, lastly, you have to offer something Superior to everything else that's out there. I don't like the new charges anymore than anyone else but for $5 a month it is cheap enough that switching to an entirely new ecosystem would probably not be on everyone's to-do list.

Sure, I actually think your last point is actually the main key. I also agree $5 a month isn't that bad, and I think Strava to a point only works as it is due to critical mass, which is hard to get. So I think the focus should be, what features do people want that Strava doesn't do (or could do a lot better). No one is likely to topple Strava, and shouldn't be an aim. More, what features would people want ?

Some of these may actually be in there, but these are what I had in mind off the top of my head (just to start off that angle of discussion if anyone was interested).

Design your own routes, and follow (with audible notifications if off course, other apps can do this eg osmaps).

Augmented reality of navigation points, or points of interest (I'm more of a trail runner!), and audible navigation.

Design your own workouts and follow them with audible notifications (like on a Garmin watch).

Have a better setup for your own running club, and have segments and workouts only available for your local club (and maybe some small subscription for features from a club could help, but free to athletes).

A beacon system which you could share privately with others, or running groups to know where others are.

Training plans, where the data would be available as to it's success.

Regarding funding, if it came to needing more, there's a few bits like certain extra features could be available only (Strava has been surviving ok without this recent clampdown). Maybe Trainers could pay to be able to contact runners (and vice versa). There's a few different routes, and obviously a balance to be had.

But the core code could be written and actually licensed out or others to use even.

There's a few different ways to skin a cat, but I'm typically just interested in having fun coding :).

Interested in any ideas others have of bits they would find useful in an app that isn't out there.

Garmin Connect > Strava.

cheaterfly wrote:

Garmin Connect > Strava.

This is why I don't really get what all the fuss is with strava. I signed up for a day (virtual 5K required it) then deleted my account a few days later as it offered nothing that Garmin connect already did for free. I use R2W as my main log and check GC if I need to see advanced data.

RobA wrote:

cheaterfly wrote:

Garmin Connect > Strava.

This is why I don't really get what all the fuss is with strava. I signed up for a day (virtual 5K required it) then deleted my account a few days later as it offered nothing that Garmin connect already did for free. I use R2W as my main log and check GC if I need to see advanced data.

Not everyone uses Garmin. A large part of Strava is a social network that is not hardware specific.

Rhodium Nights wrote:

And yet people still feel they are entitled to access this service for FREE. It is truly baffling how people can be that entitled. Strava deserves our money (those of us who use it regularly) and then some. It's a great app.

thank you. Amen to this.

Its less than a beer a month people - what is the freaking outcry on this?

yoyoyoyoyo wrote:

RobA wrote:

This is why I don't really get what all the fuss is with strava. I signed up for a day (virtual 5K required it) then deleted my account a few days later as it offered nothing that Garmin connect already did for free. I use R2W as my main log and check GC if I need to see advanced data.

Not everyone uses Garmin. A large part of Strava is a social network that is not hardware specific.

Not everyone uses Strava either, as far as I can tell. R2W, even though it is a bit dated, is still the best log out there.

runnerchef wrote:

Rhodium Nights wrote:

And yet people still feel they are entitled to access this service for FREE. It is truly baffling how people can be that entitled. Strava deserves our money (those of us who use it regularly) and then some. It's a great app.

thank you. Amen to this.

Its less than a beer a month people - what is the freaking outcry on this?

No wonder most people have no savings when you're all spending this much on beer and coffee all the time. Makes me glad I've always abhorred the taste of both.