Today’s post is gonna be a bit more on the geeky/industry side of things, though, I suspect most of you can see how this will ultimately carry itself through to more choice in consumer products down the road. Or, just simply better consumer products. Plus, the backstory and cast of characters here is surprisingly deep for a new company that’s theoretically instantiated from mid-air.
As of today, an entity known as Zone5 Ventures has instantiated itself as both a cloud platform for companies to store sports-tech data on (and they already have a long client list), as well as an incubator style investment firm. But the real kicker? They now own Today’s Plan, one of the larger online endurance sports online training platforms out there.
Today’s Plan is now a subsidiary of this newly instantiated entity, in roughly the same way that Google proper is technically a subsidiary of Alphabet, Inc. Some of this is merely moving sand into different buckets – but a lot of it with Zone5 is very much new buckets of new sand. Real humans will be getting real employment contracts that move them out of Today’s Plan and into new job roles doing new things in a new company with a different name.
Ultimately, the goal of Zone5 Ventures is to act a bit like Amazon Web Services, Google Cloud, or Microsoft Azure, but for sports-specific data. They’re aiming to handle all the file-formatting and storage mess out of it, and keeping startups and big companies alike focused on all the other features of their app or platform.
(Random Note: There’s no connection of any sort between Zone5 Ventures that this post is about, and Zone Five Software, makers of Sport Tracks and SportTracks Mobi.)
The Cloud Platform:
(Every good tech company needs a single head-tilting slide with numerous arrows and bubbles that tries to explain it all. Zone5 Ventures entrant into that competition is above.)
One of the things that almost every sports technology company underestimates initially is the challenge of the actual sports data files themselves. Sure, there are standards like .FIT files and .TCX files, but that doesn’t mean the companies follow those standards. Nor does it mean those files, companies, or standards don’t change. It can and is a nightmare for smaller platforms to try and keep up with all these tweaks, especially if their platform is looking to integrate with 3rd party services or data.
To give a very specific ‘me’ example, we have the DCR Analyzer. That’s the thing I use in my reviews to compare data sets, and thousands of you use as well. One of the biggest challenges we have is other companies doing ‘special’ things to data structures without bothering to tell anyone. Sure, you have plenty of smaller companies that just don’t understand .FIT files correctly and need help to navigate that. But, we also run into the frequency challenges of big players introducing breakages for their own internal reasons. For example last spring Garmin quietly changed how they store elevation and speed data within a .FIT file, they went from 16-bit integers to 32-bit integers. Not all units use this method, only the newer ones. Or, Suunto leaving zero and null-values in the distance accumulation data sets in their .FIT files. In our cases, these were relatively easy fixes. But these sorts of things happen constantly.
Now imagine you’re an up and coming app trying to figure it out. That’s a nightmare. I’m lucky in that the lead developer for the DCR Analyzer is also behind the FitFileTools site, and also works with Final Surge to assist them too in navigating these waters. So we tend to catch most of these quirks quickly. But if you’re trying to create the next Strava, dealing with all these nuances is impossible from a fresh start.
And that’s ultimately what Zone5 Ventures is trying to address. They’re leveraging the same backend experience we have, but in their case via Today’s Plan. They see all the same daily file formatting wonk we do, but actually have even more complicated systems to have to handle it. Which is why over the last few years Today’s Plan has actually been running more sites than just Today’s Plan. Some of them have been pretty visible – such as Stage’s site, which is back-ended on Today’s Plan. While others are less obvious, such as the Zwift Academy, or Stryd.
All of these 3rd party entities’ partnerships now fall under the Zone5 Ventures bit. In fact, here’s a partial (public) list of them below. Zone5 Ventures says that there are numerous other companies for which the relationship isn’t yet public (and may never be).
So what are they actually doing? Well, it’s essentially a collection of API (Application Programming Interfaces) endpoints into what are cloud storage pools for sports-tech specific data. They want to take any sort of fitness data and fitness sensor data and act as that normalized pool, with then further connections to 3rd party platforms and services. The idea being that if you want to create a new running app, or a cycling app, you can simply offload all of the data storage and processing bits to them. So rather than figuring out how to properly format files or anything else, you just leverage their API and you’re done. In talking with Chris Yu (one of Zone5’s founders) about it, he noted that “Just getting the basics, being able to feed data in at the top of the funnel and have that processed through to give them basic ride metrics is a huge task in and of itself”.
They’re (very specific) goal for onboarding is that a person who’s just read Coding Swift for Dummies can turn around and use their fitness cloud platform without any additional knowledge.
But it’s not just limited to app developers, in fact, they see a blank canvas when talking to other groups as well, Ben Bowley (founder of Today’s Plan) said that “The volume of inquiry we get from that kind of stuff is really quite significant, and is something that Today’s Plan is already doing. We think there’s this incredible opportunity whether it be entrepreneur or scientists.”
Which is true. The same issues that apply to a fitness startup also apply to someone trying to do research on fitness data. I see that constantly with sport-specific studies that screw up understanding sport-specific data.
Now as of right now the platform requires a bit of manual onboarding with a dedicated team. That team is currently based in Canberra, and are technically Today’s Plan employees. However, they’ll shortly be given new employment contracts and transitioned to be legit Zone5 Ventures employees where their complete time will be focused on 3rd party customers. Ultimately though, they want the entire provisioning to be completely self-service, so that a startup never has to talk to a human if they don’t want to (just like AWS or Azure).
The main blocker of that right now isn’t actually so much technology, but rather just working through the pricing pieces – which at the moment is custom from customer to customer. But their goal is that within 6-months to have super clear on-website pricing similar to what Amazon or Microsoft has for cloud services (which is largely usage driven).
They also talked about down the road offering a user-consented and opt-in anonymized sports-specific data set that would be offered to researches, some of it via grants. We see a bit of this today within the Golden Cheetah OpenData set, but looking at the scale of companies involved above, this would undoubtedly dwarf it – again, assuming users consented to their data being used for that.
Also, just like AWS or Azure, the primary goal here is a platform for other companies to build their own apps and services on – not as a data set to monetize. Zone5 says that ultimately those platforms and systems will be cordoned off from a security standpoint in the same manner as Amazon/Google/Microsoft cloud services, though, they didn’t have any specific details on how exactly that’s being achieved from a security standpoint.
They did note though that they have no issues with competitors to Today’s Plan wanting to leverage their platform (just like Apple’s iTunes leveraged Microsoft Azure and Amazon AWS platforms for years).
The Incubator Arm:
So what if you don’t have money but still need cloud platform access? Well, you can sell your soul.
Or, more alternatively – you can offload a small bit of equity.
There’s effectively two major components of Zone5 (aside from Today’s Plan), and the second piece is the venture arm, though, they noted that they are not a venture capital company “at all”. They see themselves as a bit more like an incubator arm, akin to Y-Combinator, but without giving physical space away. Instead, the Boulder-based division will give access to the cloud platform (and expertise in the sports tech realm) in exchange for equity. Technically though in addition to the cloud pillar, there’s the Incubator group, Research Group, and Portfolio group. But practically speaking Incubator and Portfolio groups are more or less the same today. And the Research group I covered under the cloud platform section.
In discussing the venture with them, one of the realities that came up is that typical Silicon Valley VC groups are generally bad at figuring out sports tech. They are easily distracted by things that have been done and died before (or are poorly thought out because they lack the knowledge of sports tech), but also inversely they’ll skip over ideas that have legit good merits because they don’t see the potential. There are surprisingly very few sports-tech specific funding groups in the space.
Interestingly though – the investment arm is open to not just entrepreneurs, but also other investors that may want to join in. However, they cautioned that they aren’t just opening the door to people with money. Today’s Plan’s Ben Bowley noted that “It’s frustrating trying to talk to people to who don’t understand this space, and it’s super hard. You end up with some in the VC space that have a personal interest in cycling, but that doesn’t always carry over to how they invest their money.”
He went on to say that the type of investors they would be interested in bringing onboard are people that have actual industry experience, rather than just an enjoyment of cycling or yoga.
And this may be as good a time as any to talk about some of the players involved here. As you’ve surmised by now, one of the founders is Ben Bowley of Today’s Plan. With him is Chris Yu of Specialized Wind Tunnel & related tech projects fame (he’s remaining at Specialized, though is splitting his time). In fact, there’s a number of Specialized-linked folks here throwing in their personal time or money. For example, Specialized’s own founder – Mike Sinyard, is personally investing funds into it. As is John Rangle, the current CFO of Specialized. There are other major investors, people I’d actually categorize as massively larger figures in the sports tech space, though, those individuals aren’t quite ready to be listed yet due to pending projects they haven’t yet announced.
It’ll be super interesting to see how this evolves. I’m often asked by startups if they know of investments that are specific to this industry, and as I noted above – the number of investors I know that actually know what they’re doing in this industry is relatively thin. And concurrently, the number of sports-tech specific services companies are also relatively thin. I could see Zone5 potentially expanding their offerings around the services side of things – potentially to include hardware/sensor consulting/advisement as well (similar to what NPE does behind the scenes for a number of major customers and gear you’ve very likely got in your pain cave today and don’t even realize it). All areas that new (and old) companies struggle with.
And of course, I’m interested to see how quickly Zone5 can shift from manual onboarding of customers into the cloud to a fully automated and self-service platform, which is key to getting 1-3 person startups onboard that don’t really want to deal with lots of administrative or contractual overhead. If they can hit their 6-month timeframe, that’d be pretty impressive.
In any case, hope ya found this interesting, and thanks for reading!
Super interesting. I assume they offered you a job, and I assume you declined. 😉 I mean, a venture like this needs somebody like you.
I think what they’re doing is pretty darn fascinating, but I’m pretty happy doing what I’m doing.
Very interesting.. If they can get some traction for the concept of building a true cross vendor repository, especially one where the athlete (not the companies) can dictate how their data is shared and used, it could be a real thorn in the side of Strava. The power play with Relive and others over the ability to offer capabilities that might compete with each other would be much less likely if the rider/runner was in control.
I asked about that (the Strava side of things), and they said they aren’t really targeting commercial data set usage or monetization, versus the core research target. They aren’t specifically opposed to it with the right user-driven controls in place, but it wasn’t what they are focused on now.
Strava’s main risk is that investors stop giving them money and that users continue to realize Strava isn’t providing new feature value for their continued monthly/annual membership fees (which is the entire premise of subscription based servers: ever green features).
Just canceled paid strava account after three years. Stunning how little they have done with this massive head start. Have they added any noteworthy features in the last three years? Strava shows the power of of network effects but strava is testing it’s limits. I would love it if this back end platform could be a start to some true innovation in the space. If they could figure out a way to do some sort of shared public segments it would be amazing – it would put a nail in strava’s coffin. A distributed segments database would be amazing… wonder if blockchain could address this.
Long gone from Strava. Just got an update notice from a 5 year old feature request they still have not implemented.
“It would be great if I could see the time and average pace for manually created laps in the mobile app, such as when we do interval training. Ie for when we do 400m or mile repeats.”
Think this was requested alot during the last years. Standing on track not able to view your own laps was most anyoing.
Ray – Have you heard anything new about the Strava/Relive stuff? Any chance they work something out? I’m a long time fan and paid user of Strava but just hate this kind of dumb stuff.
Why anyone ever gives money to Strava is a mystery to me. I was a premium member around 2014-2016 but soon noticed that there was not, and there continues to not be, any feature development. Strava created a massive rod for their own back with segments and spend all the money they have keeping the processing necessary for segments working.
On the other hand Garmin Connect has steadily developed. I get a lot from their training metrics.
Here’s the thing – the only thing I ask around here is that the comments section doesn’t become the wasteland that most news sites comments sections are. If you think it’s wasted money, than actually provide an opinion – heck, even just a name. Not just two words that are meaningless. Otherwise it’s just junk/spam.
FWIW, DCR comments are one of the few comments sections that I bother to read. While I do not always agree with everyone’s, or even your, opinions (heck sometimes I don’t even understand them) I always appreciate the thoughtful and respectful tone.
Validic (link to validic.com), also a Durham NC company, has been doing something similar for about 10 years. They originally launched more in the fitness area but pivoted (slightly) to health and found more success there – it’s a fully featured PAAS with data storage, API adapters, and service consumers from major health + fitness data sources.
I think the big struggle here is finding the market space between… at the low end – using existing dev ops infrastructure plus some open source libraries to read data + interact with APIs, to the high end which is essentially using an existing platform as a black box + a veneer of rebranding.
This could be a great tool for fit startups that want to prototype or get a rapid MVP up for a pitch to venture funding. But… in the big scheme of things, if you’re going to launch a sports fitness platform that you plan to scale to Strava size, or if you want something with longevity – the effort this saves you isn’t **that** significant… And the downside – you’re committed to a proprietary set of tools whose price is likely based on volume, and you can’t just search Stack Exchange to fix problems.
Yeah, I think it’s a perpetual balance for tech startups between trying to balance designing a system that can scale to tens of millions of users someday from the start, versus designing one to prove commercial viability of the product.
I don’t think there’s any right answer, though, I think the data processing aspects of this would probably quite easily scale even at Strava’s site. After all, Today’s Plan does a better job of data handling than Strava does for certain types of data (like swimming).
Why only MVP or prototype? Many of the smaller hardware vendors struggle to provide a compelling cloud backend for their gear. Those might very well prefer to rent a working whitelabel solution from a supplier over hiring a small team that will never get beyond the barest basics before getting locked into the bugs-caused-by-bugfixes cycle. Their typical customer won’t be a miniature Strava, it will be a miniature Garmin.
I think it would be super interesting if the likes of First Beat would licence their algorithms to a platform like this. It would then be easier for the smaller sports tech companies (anyone remember Timex) to start to exploit them on a pay per use basis. This may even allow customers to take advantage on new metrics as First Beat releases them if they could be a monetisation model were the customer pays for additional metrics rather than the case today where the like of Garmin would have to licence a feature for a back catalogue of products
I am most interested by the technologies! As someone working on OpenSource Cloud Techs (ie: Cloud Native foundation/Kubernetes techs) I am most interested by how locked down those startups are going to be when choosing a ready to MVP package!
If advanced First Beat metrics are made available through a cloud service like this, what’s that do to Garmin (who love to move customers from a ForeRunner 245 to a 945 just for more Firstbeat metrics, basically)
If you could easily and cheaply get as many metrics as you want, hardware companies couldn’t upsell as much as they want!!!?
Are they going to have their own app, or is that being left up to partners/ third parties?
Right now, i’ve got Data strewn across several apps, and this seems like a step in the right “central health/fitness data repository” direction. Just wondering what the best method will be for organizing/ viewing everything and getting recommendations based on that info.
No plan at the moment to provide a consumer-facing app that I’m aware of. This is more about the behind the scenes data repository.
tapiriik.com helps — it synchronises, without duplication, across platforms.
and the code is opensource !
In all honesty, this sounds like an amazing idea and I’m glad that you posted about it.
First, it was enlighting to read that sites like Stryd or Zwift actually send the data to a third party, and while that can be controversial without the right controls, it’s actually interesting from a user standpoint for data analysis.
I have seen some inconsistencies when comparing data from sites of my own training, so having something centralized helps alleviate that problem and also help improve the standardization in the long term, cause even if they don’t dictate it (which they won’t), it will help the right and quick adoption to the modifications that the providers execute so we can eventually compare apples to apples.
Thanks, Ray for this article! Love the geekiness of your blog and your articles.
What impact might this have on Relive or another App like that that used to have partnerships (with Strava) but were cut out. It would be interesting if the user had control of their data through the (Zone5) API to share it with whomever or whatever they wanted without having to fear that one App would suddenly cut off access to another App.
I don’t think any specific impact today, it’s too abstracted right now to impact that relationship.
I’d certainly like to hear/see more about this. As may be evident from other posts, I’m a bit of a stats junkie (when recorded accurately, okay, I’m a stats hard-ass, LOL) Anyhow… on my own I keep a spreadsheet of data from back when the Polar watches (and CatEye computers, etc) didn’t sync with sensors, etc. It’s a very wide spreadsheet, from years of accumulated “features” that have appeared or disappeared across the various devices (and multi-tabbed, for different bikes for instance that have come and gone for example so I can keep track of each bikes mileage,) so I have a “Running” tab, various biking tabs, and (should have) a swimming tab, for example (so far I let Flow manage swimming as the v800 is my only swim tracker thus far).
I could easily turn that into a database (and probably should have in retrospect at some point probably going to now that I’ve thought more about it) but as mentioned in the article, keeping up with specification changes, is easier when it’s human readable values on a spreadsheet, versus a machine parsing into a db “typed” field. If I put in 60%, or 60 in a “%” column, I can read either easily enough and sort out what I meant. If vendor A records 60% as “60” and vendor B records it as “0.6” and vendor c records it as 39322 (60% of a 16-bit value), that’s where the fun of translation begins.
Obviously a data abstraction layer for each generation/vendor/type of product has to be involved, as Zone5 said, getting it “into the funnel” is the main challenge, how “I/they” store the value on the backside ultimately doesn’t matter so long as the field is suitably wide enough to cover all variations with an appropriate modifier type applied, and may (hopefully if designed right) never need to change (but if it does, a generational ID modifier can handle that too,) but how they PROVIDE the value to the data integration could be any of those or other options like above. (NMEA, for boaters out there, is similar, a lot of tweaked/customized fields even though it’s an official “standard”.)
I was going to go into my experience of exporting data from Polar Personal Trainer (PPT, long story short, old data from one of my watches, won’t auto-migrate into Flow, due to customized HR zones (3) not matching the Flow-verse (5) zone it seems. There is SO much wrong there IMO (including the output being xml, but I sorted past that pretty easily). There isn’t any reason a junior DBA couldn’t have coded the “broken” data mapping to transfer to Flow in probably a day’s efforts, either by dropping data that can’t be matched, or by translation of data that can map over. Instead though, they just don’t do it (in this case, I understand you can’t match 3 accumulated data zones, to 5, as you have no idea where the “breaks” are in the zones.) But it’s up to the end-user (95% of who will have no idea how to even map/manipulate the XML output) to gather and handle moving their data in a meaningful fashion.
I’m pretty sure most users would have been okay with “we can’t get all your data over, but we can pull StartDate, StartTime, min/max/avg HR, total distance, total time, min/max/avg cadence, and min/max/avg power (etc/whatever). Hope you understand, and if you want to save the rest of your data, here is how to export those details:” Some of those things simply don’t change and aren’t device-feature-specific.
The other point to that, the data stored in PPT for “distance” even though it displays as “Miles” (user-selection) is actually stored in meters. Something I had to sort out and convert, only took a minute to sort out as a data junkie and with a database conversion history I expected it, but for “Joe-User” “okay, I have this random gobbledy-gook XML file I can barely comprehend, I see something that looks like distance, but it’s just a number, 29,141… what the heck does that mean???” (18.1 miles, FYI).
Hopefully if the “fitness tracking world” at large could all move to something like this service, then standardization, and long-term retention of meaningful data could become much more reliable. Or a massive standardization of what and how to record (NMEA again for example). .fit, .gpx, .tcx, all are “standards”… but all have variations, and with new recording devices/methods (running pods for example) may require on-going extensions. (Windspeed, from Stryd (or that bar-mount aero sensor, for example) maybe a future output value to record, if not today?)
In unrelated news, they’ve got open career positions on their website…
> …a massive standardization of what and how to record …
What about JSON, link to en.wikipedia.org as a “man-in-the-middle” ?
You would think there would be major potential here for AI development with such as massive data set.
Especially to detect patterns in training that could be consistent with the use of doping products. In theory, having a grand unified sports data platform would be almost exactly the same as the biological passport based on watts, in my opinion.
Does this affect training Peaks at all?
Directly today, no. Indirectly down the road, yes.
It’ll make it easier for up and coming TrainingPeaks competitors to get off the ground. And, it gives Today’s Plan more data and experiance to work with – which will ultimately give them an advantage over Training Peaks. But that’s mostly pretty subtle, and given Training Peaks is so well connected industry-wise, it’s not like someone would skip over telling them about something.
Mulesoft Works with APIs
Maybe they can help them with their experience
Are there potential issues here with personal data, GDPR etc. Due to the increased personal data collected by these apps and devices? The scale of opportunity here is huge – wish I had some money to invest!
Interesting to see developments in this space! What I’d be interested in is how much of this is consumer facing vs. a cloud platform as you describe. In short:
– End-users directly sync their data to Zone5 (Zone5 is the data controller)
– Zone5 acts as a centralized hub where developers can access data via an API (developer application is a separate data controller)
– Zone5 acts as a gatekeeper to the user data, charging developers for access to the data
– End-users sync their data to the application built on top of Zone5 (application is the data controller, zone5 is a data processor who processes data on behalf of the data controller)
– Zone5 provides the building blocks to handle different APIs, file formats, etc.
– Data can still be stored at Zone5 to simplify application development, but is scoped specifically per application. Zone5 can’t / won’t access it similar to how AWS won’t interpret any of the data in S3, but only provides the storage layer.
– Zone5 charges for storage and platform use
Differences might seem subtle but pretty fundamental if the goal is to build an independent platform
You nailed it.
Do they want to be the Apple App Store for Sports -or- AWS Sports ?
Definitely not an Apple App Store for sports.
I think the best blend they’re aiming for is AWS for Sports, but with the normalization of something like Healthkit (kinda, sorta).
This is relatively interesting, there are some significant concerns associated with this, that if done right would be a major step forward.
My main concern is who is their customer? and who owns the data are they storing?
The problems in this space start and end with the tensions between those two questions and the interests of the consumer.
I think in some ways the answers to those questions are already within the customers they have today. Meaning, Zwift, Stryd, etc…
So at present (for the foreseeable future), their customers are B2B in nature (plus researchers, though some of those would be grants, so ‘free’ customers). The data sets are isolated, in the same manner that Team Ineos (and previously Team Sky) ran isolated servers for their data. The data is governed by the agreements between each respective company/entity and its users (in other words, defer to Zwift ToS/etc…).
It sounds like longer term though Zone5 wants to be able to go to companies and say ‘Hey, can you add an opt-in option for researches to access data sets’ or say ‘Hey, we have this updated/new API for accessing data in a different way’.
Like most PaaS entities, there’s going to be a case of legacy/unique customers for quite some time. I suspect that some of the logos picture have super unique situations (from a dev standpoint), whereas others are more vanilla/cookie cuter.
I think the good news here for consumers is that in talking to Zone5, they didn’t seem super interested in monetizing the end-data itself. Meaning, I asked if their goal was to then sell things like annon (with consent) device data sets to companies for market research/etc, and they said that wasn’t really their interest. They wanted to focus more on the academic/research angle for those anon/consented data sets.
The good news here is that they’ve already got customers with real-world deployments and users in place. So we kinda know the answers to these questions, at least for the immediate future.
It looks to me like they are just a PaaS with some pre-canned sports data capabilities.
The promise of open data sharing and/oe services that exploit their data collection are hollow unless backed up legally in the ToS -or- a direct relationship with the consumer.
If they can sort that then this could be a brilliant development, but without it they’re just AWS Sports.
I have long wanted a way to use XaaS for free (as in freedom) solutions, but have been constrained by the need to pay on behalf of my users. The inability of the XaaS providers to let me pass the costs of service on to consumers in a way that is compelling and removes commercial risk from me as the solution provider has always been a deal breaker.
Will be watching carefully.
Any word if Xert is lined up for using this?
As someone who has worked for an endurance sports fitness tracking platform for many years, I just wanted to say that there isn’t any shortage of smart people and startups trying to create new apps and services in this space. I find new “competitors” every week, sometimes multiple times a day. I’ve even found new ones that work out of the same building we do on more than one occasion. It’s true that keeping up with little changes that break things is an ongoing challenge, but from my perspective there are far greater challenges that must be overcome to succeed longterm, and they’re all things that readymade services will never be able to provide.
If someone could take the data from millions of users and actually see what works as far as making one faster…I’d pay a good bit of money for that. Strava premium takes a stab at it, but it’s basically train harder and you’ll be faster. While that’s likely true, I’m sure there are smart and dumb ways to approach that philosophy.
Problem would be the accuracy of the data. I can’t be the only one whose measured HR starts spiking at 190 when running at an easy pace or drops to 90 in the middle of an interval (no, not my real HR). Or those times I just forget my monitor, the twice/week weight workouts I don’t record, etc.
Any news from this? With the latest strava change there might be a good time now…too many services depend on strava for their data sync….
Ray, what do you think?