How to analyze products
An example: comparing Runkeeper, Endomondo, and Strava run tracking applications; TLDR
In this post, I compare three popular run-tracking products: Runkeeper, Endomondo, and Strava. I compare the product from three perspectives: usefulness, usability, and business. Of course, these perspectives are not independent, the usefulness for a runner and usability of a product influences its business future, and the business model of a product can have a considerable impact on its usefulness and usability. Also, I focused on the features available with the free versions of the apps, because most of the users meet with these features first and make the decision of spending money based on their experience using the free version.
In 2009, with a colleague, we developed an Android app, CoolRunner. From the business perspective, the app was a huge failure but both of us (having B2B background) learned a lot about the world of consumer products. In this post, I also rely on the lessons we have learned from that failure.
Why is the analysis consider only runners, you may ask. First of all, I wanted to limit the size of the post, but even so, it is too long to read. Secondly, the market for running itself is big and growing enough to support several apps. Thirdly, and least importantly, I run way more than ride, do yoga, or swim.
Market size
The market size is determined by the number of runners and smartphone penetration worldwide. In 2016, in the United States alone about 47.4 million Americans participated in running/jogging and trail running, but less than 2 million finished a race, and 0.5 million runners finished a marathon. I could not get details worldwide statistics, but based a post on runrepeat.com, I estimate the number of people participated in running worldwide to roughly 250 million, and the number of marathon finishers to 2.5 million (based on the NYC Marathon statistics, 1–2 percent of the participants could not finish, therefore I estimate the number of participants to be roughly the same as the number of finishers).
Smartphone penetration in the developed world is between 40–90%. In the market size estimation, I assume that at least 70% of the runners have a smartphone, which gives the following worldwide run tracking market: 175 million runners and 1.75 million marathoners.
User groups
I intentionally don’t use personas, because, given the available data, I could not provide well-founded persona descriptions. Instead, I define high-level groups of runners, which could serve as a starting point of persona definitions, and also can be used to analyze the products.
Definition: User groups and their needs
On high-level, there are two main disjoint (at least at a given time) groups of runners: fitness runners and racers.
- Fitness runners want to stay fit and healthy and may want to lose some weight. They may have run a race, but they are not hooked on racing. They also tend to skip some workouts due to various reasons like bad weather, or family or work obligations.
- Racers regularly attend races (say, at least once every year), they may run marathons or half marathons. Running is basically one of their hobbies, and they strive to improve and get better by well-designed workouts. They also conscientiously prepare for their upcoming race: they analyze the track, prepare for the weather, determine their race pace, and discuss the specifics of a race with fellow participants.
Based on the latest statistics we know that all over the world there are roughly 175 million fitness runners with a smartphone, and the number of racers is roughly two orders of magnitude less.
Even though the primary runner groups are disjoint, they share some common needs:
- They don’t want to get injured.
- Availability and accuracy of data (e.g speed, distance, ascent/descent, location, duration).
- Don’t want to spend much time to learn the product: ease of use and aesthetics.
- Minimize the interaction with the mobile app during the run.
- They want to save battery.
I also defined two secondary groups of runners. Both fitness runners and racers can belong to these secondary groups, and a runner may belong to both secondary groups during a short time period:
- city runners usually have to stop sometimes, typically due to a traffic light, and they may run among buildings that mess with the GPS signal, while
- continuous runners usually can complete a run without having to stop due to external circumstances (like having to stop to cross a road), and their environment rarely disturb the GPS signal.
High-level use cases
In order to address their needs both fitness runners and racers want to record their runs, and get feedback during a run (the details of feedback may be different in the case of the different groups). They also want to check the data of a previous run (the details may also differ per user group), and get aggregated statistics over time.
Fitness runners are more interested in
- checking the weight loss as a result of a run or during a given time period (e.g. per week, month),
- specify and a distance and/or time goal and get notified when the goal is reached,
- sharing a run on social media,
- getting motivated.
Racers appreciate the use cases of fitness runners, but they more interested in
- setting the type of the training (e.g. interval),
- segment workouts according to relevant parts such as warm-up, cool down, fast and slow intervals,
- analyze data, including checking trends (e.g. personal bests),
- looking up certain workouts based on their relevant parameters,
- plan (and re-plan) workouts,
- see the details of an upcoming race (e.g. route, elevation, typical weather), and
- discuss race with other participants,
- compare their race performance with others (to see what they did right and wrong during a race),
- get race forecast (e.g. pace),
- get advice how to improve.
City runners want the app auto-pause recording when they stop at traffic light.
High-level strategic considerations
As a rule of thumb, people don’t want to buy software, but people buy products what they trust and deem important in their life. Trust takes time to build and requires investment from the user and the company which develops the product.
People often rely more on their feelings and social relationships when they make decisions. Therefore, unless a product provides an essential service (such as a water-pipe), it shall be both aesthetical and fun/easy to use (note that water-pipes can also look great). Besides, it also helps to decide if we know what products do our friends/acquaintances or famous people use, and social networks may have a significant role in making this information available. Note that trust can also be built faster if the product relies on the social relationships.
The more time the user invests in the product, the more probable is that the user gets hooked on that product, and users, who are hooked on a product will return and interact with it regularly. This, in case of a run tracking application, is a win-win relationship: more interactions mean more physical activities for the user, and more opportunities to sell to the company (e.g. sell ads, features, services). Therefore, means of motivation are extremely important, especially in the case of fitness runners.
Not only the sheer number of fitness runners make them the more important user group. Many racers start as fitness runners and get hooked on running afterward.
Collecting as much data as possible, and showing the right data is essential. It is relatively cheap to display raw data and can be impressive for all users, while the real value is in the insights. Although racers love to dive deep into their data, good insights have high value for them.
Data can be utilized outside the immediate borders of the product if it provides extension points / API for developers. Thus not only the data can be utilized, but also the innovative power of other companies which may provide further useful or fun to use services/products. These services/products may not be limited to the domain of running, they can range from healthcare to games.
Finally, the opposing needs of the main user groups must be considered. Fitness runners long for simplicity and fun, while racers want the abundance of data and insights, and the product either should focus exclusively to one of the groups or should be structured so that both needs are satisfied in a usable way.
Key metrics
In this section, we take a closer look at the metrics that describe the behavior of the customers of a product, and therefore fundamentally determine the success of that product.
Acquisition
There can be several potential channels of user acquisition. Because there are two orders of magnitude more fitness runners than racers, and racers usually start as fitness runners, and they may already use an app before their first race, I consider the fitness runners as the more important group of the two in the context of user acquisition.
In case of fitness runners, I believe that word of mouth and social networks, especially among friends and fellow runners, are the most important acquisition channels. Regardless of the channels, the product shall promise that users will stay fit, exercise regularly, and lose weight (if they want to). The apps and websites must be visually appealing, which, in case of fitness runners, chiefly means focusing on the basics and ease of use, while in the case of racers requires offering a free function which gives some non-trivial insight or piece of information.
In case of racers, advertising an app as the official race tracking software might seem to be a good idea, but I think that those, who have already participated a race have also already used an app and invested into it a considerable time. Therefore, unless the offer includes the usage of pro features (including training plans) for an extended time period (say, a year), not many users will switch to the advertised app.
Activation
The first impression of the runner about the app is crucial: the activation shall be super-smooth, the users shall be able to install the app, register, have the first run (start, pause, stop & save), and see the results with the least effort. Besides, the benefits of using the product should be obvious instantaneously for both main user groups:
- fitness runners should have fun, get encouragement, and see the positive impact of the run to their health, while
- racers should get non-obvious insights about the run and/or and/or a possible race.
Retention and engagement
Motivating the users to run is the key to achieve high retention. The more the user interacts with the app, the better it is for the company and the user himself, therefore high retention is beneficial for both of them. Engagement with the product does not stop at the end of a workout. Users may analyze the data, interact with each other, get nutrition and training advice, etc.
To motivate users the product shall generate positive emotions associated with the efforts of the user. Products can provide extrinsic motivation via trophies and rewards, and foster intrinsic motivation by allowing to set short- and long-term goals and highlight when they are achieved. Achieved short-term goals (e.g. run a 20k every week) give positive feedback regularly, while long-term goals motivate the runner over years (qualify for Boston marathon). Experiencing a sense of progress towards the short and long-term goals is a huge motivational force, and the product, in order to boost retention, shall facilitate it. Also, if a goal is visible to everyone, the user will be more committed to achieving that.
It is the product’s best interest to help runners to achieve their goals, even free of charge. It can be achieved in many ways such as giving nutritional advice, monitoring their sleep patterns and warning when the user has not slept enough, showing videos about running techniques, giving training advice, recommend people to follow because they achieved a similar goal or they are professionals who are good to listen to, etc.
Offering an achievable goal with a free training plan can also have a positive impact on retention, and helps to build trust, which is imperative to get the user to pay for a service or feature. For example, if a user can run 10k in 44 minutes, and the product asks if he wants to run 10k within 40 minutes within 8 weeks, the user can easily say yes, run more, interact with the product more, and if he really achieved this goal, he may also buy a sub-3-hour marathon training plan (because the 10k plan worked).
It is harder to stay home on a rainy day or when the user is tired if he has a running partner who is waiting for him outside. The product can help the runner to find such partner or partners by connecting runners. Although actually running with a partner is the best to start training even when the circumstances are not ideal, but knowing that people are out there in the cold or rain also helps.
Injury prevention helps to preserve retention. Increasing the weekly running mileage too quickly, running in an old shoe, running in too many races, etc. all can lead to injuries which decrease the usage of the product.
Although I emphasized the importance of running in retention, I also referred to such services of a product, which are not bound to a specific workout. These are pieces of advice, videos about running techniques, etc. Race reports are also interesting reads for fitness runners and races and can increase retention and engagement beyond workouts (e.g. on rest days). Consequently, if a product has a conscientious social network strategy, and aim to serve as a social network for runners, it can boost retention and engagement way beyond of simple run tracking apps.
External triggers can also prove to be useful to direct the user to the product and increase usage. Such triggers can notify the user about various events, for example, when a friend finished a 10k run, improved his 5k PB, a new post appeared about nutrition during a marathon, or a new report appeared a race he finished a few days ago.
Needless to say that the products shall not annoy the user in any phases, but especially during the activation. The user will immediately decide to uninstall the app and switch to another product, if he has to interact with the product in an extremely unusual way, or has to confirm actions unnecessarily, or if he faces with messy screens (including useless charts and data), if the app pops up notifications every now and then, shows ads everywhere that ruin the experience or terrorizes him with subscription offers, or if the app loses his workout.
Monetization
Retrospectively, we have done everything to fail with CoolRunner. The biggest mistake that we have done is that in the beginning, the app did not have a free version, and consequently, due to many reasons (e.g. lack of trust; availability of simpler, but free apps), our user base too small for advertisers. So we made little or no money from selling the app itself, and after we made it free, we had so small user base that our ad revenue was also depressingly low. Therefore, we have learned, having lots of happy users is the pre-condition of financial success in this business.
Revenue from ads can exceed other revenue sources by several orders of magnitude. Ads can appear as banners in the mobile app or the website, or sponsored challenges, events, posts, videos, rewards (e.g. coupons) which relate to running, and potentially interesting for the users. Again, ads shall not spoil user experience and shall relate to running, sports, or healthy lifestyle, otherwise, they will hurt retention and lead to high churn.
The strictly advertising-based revenue stream which relies on banners, however, is threatened by several things. First of all, in case of the mobile app, even though a workout may last for several hours, during that the user has little or no interaction with the app itself: he rarely if ever looks at the screen of the phone, or touches any control (and surely won’t tap on an ad intentionally). Secondly, the training analysis lasts for a few minutes, which gives little time to show ads. Thirdly, even though ads can appear on the website, many users have ad-blockers. Sponsorship based ads, however, can overcome these threats, since the users can regard them as an integral part of the product (e.g. a reward of a challenge).
Users may be offered to purchase the ad-free version of the app. In theory, this can also threaten the mobile ad revenue, but with careful pricing strategy, this threat can be mitigated.
Although in our experience at least two orders of magnitude fewer users are willing to spend money on a run tracker product than the number of users of the free version, in case of a well-established professional service, revenue from subscription and in-product purchase can compete with ad-based revenues. It requires a detailed analysis what features and services shall be sold, and in what model, but as a rule of thumb, a service which has significant value for the user but loses its value over time (e.g. training advice — what to do next — based on previous runs) or used frequently can be made available as part of a subscription, while such features and items that have reproducible (e.g. 12 weeks marathon training plan) value or used rarely can be made available permanently after an in-app-purchase.
Note that it is important to carefully consider the pricing of features, items, services which boost retention and engagement.
Finally, third-party developers can also be the source of a revenue stream. Access to data and services can be charged based on different business models (e.g. yearly subscription, pay per amounts of data accessed, revenue share). Again, similarly to the regular users, it is worth to follow the freemium model: 3pp developers may get access to most parts of the APIs and limited amounts of data free of charge.
Referral
The social network of the runner has a significant role in selecting a product (see Acquisition), however, people don’t have detailed knowledge of the hobbies of their acquaintances (for example, I just learned that one of my ex-colleagues is an ultra runner). The ability to share training results on social media (e.g. Instagram, Facebook, Twitter), can help the viral growth of the product, but it can also give positive feedback to the user after he gathered kudos from his friends.
Apps offer to invite friends, which is somewhat awkward to me, because unless a friend is just looking for a run-tracking app, he will never even click on the link attached to the invite, and he may even be annoyed by it. However, if a user shares his honest and positive experience about a sporting achievement such as a successful training, losing weight/burning calories, or finishing a race with his PB, and that achievement resonates with his friend’s goal/desire, he may give a try to that product.
Churn
I believe that users should stick with a product or service because they are satisfied with it, and not because they cannot migrate to a better one. This said, after a user invested a significant amount of time to a run tracking product, collected a huge pile of training data, he will not consider switching to another product even if a better alternative appears on the market. But if the needs of the user change significantly, say he becomes a racer after years of being a fitness runner, he may consider switching to another product, which fits better his needs.
Therefore, in order to keep existing users, it is not enough to satisfy their immediate needs, but it is also important to address their future needs at least on that level that makes the migration to another product relatively costly. (Note that although tapiriik.com allows users to synchronize data among various tracking products, I assume that it is known by a negligible proportion of all activity tracking product users, therefore I consider its role insignificant in churn.) Concretely, if a fitness runner wants to run a half-marathon, and another run tracking application may fit his newborn racer identity better, his app, which might be tuned for fitness runners, should offer at least such features/services that would make them too costly compared to the benefits. Again, given the two orders of magnitude difference in the number of fitness runners and racers it might not be worth the effort to develop such features that will be used by a marginal number of users, but given the willingness to invest into their improvements, with good pricing strategy, developing features for racers can be made profitable.
On the other side of the story, users can be attracted from a concurrent product, if their needs are significantly better served. Consequently, apps have to decide which user group they want to serve primarily.
A note on the impact of the social networking features on the metrics
Social networking features became integral parts of the run tracking applications. You can register with Facebook, find your friends, like each other’s activities, share your activities, and do many more things. The way a product connects its users and allows them to build communities can have significant impacts on the user behavior, and the corresponding metrics.
Applications
(Disclaimer: I am using Endomondo for years. Before that, I was using my own app, CoolRunner. Yes, I know, it is a shame…)
Runkeeper
Runkeeper states that “our mission is simple: we want to get the whole world running. That doesn’t mean you have to do 26.2 miles, but working on a product that improves people’s lives should excite you. From race training to regular old endorphins, we’re helping the world get healthy the best way we know how: together.”
First of all, there is no mention of social features, and interaction between friends/runs (yes, they write “together”, but that could refer to Runkeeper and its users; if they wanted to make it clear that they facilitate social interactions, they would mention it like the two other products).
Their main goal is to the world get healthy by getting the whole world running (i.e. running is not the goal). Consequently, based on their mission statement, Runkeeper seems to focus on fitness runners.
Endomondo
Their mission “is to motivate people to get and stay active. We do this by making it more engaging, more social and more fun to exercise.
…
We believe fitness should be fun and that social interaction is the key to achieving this. Whether it is a real-time pep talk from your significant other, an inspiring comment from your best friend or colleague, or likes from other friends in your social circle, the positive reinforcement will help you go that extra mile. That’s why the app is designed to make it easy for you to connect with your friends, and we recommend that you invite at least a couple of your close friends to try it out.”
Based on this mission statement, Endomondo is more for fitness runners: their main goals are motivation and having fun. The social interactions are the means to achieve these goals.
Strava
According to the mission statement of Strava, it “is the Swedish word for “strive”, which epitomizes who we are and what we do: If you’re striving to improve, no matter your goals or ability, you’re one of us.
Strava is the social network for athletes. We’re a global community of millions of runners, cyclists and triathletes, united by the camaraderie of sport. Our website and mobile apps bring athletes together from all walks of life and inspire them to unlock their potential — both as individuals and as communities. From Olympians to weekend warriors, we’re out there on the road and trail, all over the world, day after day.”
So in my interpretation, Strava aims to help its users to improve and unlock their potential, and it is primarily a social network for athletes and not a run tracking application. From another perspective, its social network features aim to help runners to improve, as opposed to having fun or “bringing the world closer together”.
Consequently, barely reading the mission statement, Strava is more for racers than the fitness runners.
Comparison from user perspective
I took my old iPhone 5s, put myself in the (running) shoes of the fitness runner and the racer, and I run with all three free apps during a 10 weeks period. After the runs, I synchronized the data to the other apps (I used tapiriik.com for that), and analyzed the workouts using the websites and the apps as well. As a result, in the last few weeks, I was able to get the hang of them, and if I missed something though, I consider that as a usability issue. I also trained with all three apps tracking the run so that I could exclude issues resulted from data synchronization.
Here are the results:
Common concerns
Injury prevention is very limited in all three free apps, which besides the potential health issues also hurts retention.
The inaccuracy of running stats in the city can be traced back to inaccurate raw GPS data due to buildings. While I could feel a certain level of empathy remembering the time when we were struggling with GPS data, if we could solve it in 2009, the should be able to do it in 2017! If we take a look at the route maps it will be obvious that the locations scatter so much that the actual distance covered during a training is somewhat shorter, therefore the speed must be is lower. (Runkeeper tries to solve this issue with the ”fix GPS” function, but I could not see any detectable improvement on the map after I pressed the button, and I also wonder why do the users have to request it after a workout if it is well-known that poor GPS affects accuracy, and why cannot it be done during a workout.)
In case of Strava and Runkeeper elevation gain was especially far from accurate. Even if you run on a surface which is flat like a pancake, raw GPS height data will scatter around the actual height: the GPS will give the APP sometimes smaller, and sometimes higher value than the actual altitude. Strava and Runkeeper do not consider it, and even though I ran on an almost perfectly flat surface for 16k, it summed all the inaccuracies of the GPS and showed me that I climbed 137 and 113 meters, respectively. I am very bad at climbing, and I would have felt that for sure! In contrast, Endomondo reported 20 meters (way closer to reality).
The Strava app, on the other hand, was the champion of auto-pausing, which is absolutely necessary to have accurate speed stats as a city runner. When I stopped at a traffic light, Strava stopped the timer, while the other two apps often took their time, and showed on their screen slower and slower current paces (I was wondering, how come is it possible that I am standing at 15:00+/km pace) until they finally stopped. This leads to longer recorded distances and slower average paces. (In CoolRunner, in 2009, after we realized that GPS raw speed data is also inaccurate, we both were using the accelerometer data and a lower speed limit to stop the timer immediately when the runner stops. The “wonder” of dogfooding…)
Being a long-time Endomondo user, aesthetics, and ease of use were hard to judge to me. There are some objective issues with the controls on the Strava web: it is often unclear, what is a clickable link, and what is actually a text. Besides, the web pages contain clickable items here and there, and the user really has to spend some time to explore the pages. The Strava app design reflects the “social network for athletes” mission, consequently, it is not the track recording, but the activity feed is the default screen. If you want to record a training, you need to switch to another tab and leave the feed. Seemingly, from a developer perspective, this is an obvious solution, but the mental model of users about social network feeds does not fit this design (i.e. we never go to another tab to post to our Facebook page).
Based on its look and feel, my first impression about the Runkeeper app and web was that it was built for younger runners. But when you compare the activity feeds of the three apps, Runkeeper feed items are so poor that I doubt that they’d be appealing to a runner in his early twenties: there is not may, no summary, etc. Besides, just like in case of Strava, when you open an activity on the web, you cannot add a description: you have to find and select a drop-down (non-trivial), click edit, find the description input box, add the text, and press save.
Awkwardness is everywhere in Runkeeper. You can stop and save a training with four taps, while you can do it with only two taps in Strava and Endomondo. Somehow Runkeeper named my Friday run “Péntek Run” (péntek is Hungarian for Friday), however, I use English on my phone and in the product (web and app), and when I want to edit the activity on the web, I can change everything but the title of the run (e.g. snap the track to the road, which is actually smart, but why isn’t it done automatically; I don’t zig-zag among cars as raw GPS data suggests!). But it is not a big deal since the title is not used in the activity feed, so I don’t really get what is it actually for.
Endomondo is also far from perfect. One of the most annoying design flaws relates to the “status” and “personal notes” of a run. What is the difference between them? Both are personal textual descriptions of the given training, so why do we have both? Before that, we had a title and a description, and I understood what are they, but these things are a mystery to me. We still have a title, a big bold text above the light gray starting date & time, but it is not visually obvious that (a) you can click on it to (b) edit the title. Also, there are few things more annoying on a web page than single-line editors for potentially long text. Yes, users will eventually learn that the single line “status” and “personal notes” fields will expand when text reaches the end of the line, but when you look at it, it is far from obvious. Finally, with light-gray text on a white background, you have to set the contrast and brightness of your screen properly and you should not have problems with your eyesight to find those fields on your screen.
Fitness runner concerns
Two of the primary needs of fitness runners are not addressed directly by the products. How do I stay fit and healthy? You will not get nutrition and training advice from the apps. They have a blog that you can check (if you find the link), but users interface with the apps, the websites, and if anything, they check their activity feed.
Similarly, if our fitness runner wants to lose weight, he gets some help, but none of the apps address the problem directly. The users get the calories they burned during a single run or a time period, but it is unclear, how does that translate to actual kilos or pounds. Of course, weight loss is more complicated than showing the mini cupcake equivalent of the calories burnt during an exercise, metabolic rate and calorie intake are also part of the equation. With Endomondo the user can follow his weight and calories burnt, and with Runkeeper the user can follow his weight and body fat over time, but without the calorie intake, the chart is incomplete. On the other hand, all three apps integrate with MyFitnessPal, however, the user has to know about the app.
The products use various gamification techniques to motivate the users to run. Interestingly, goal setting is a premium feature in Strava, while Runkeeper offers it for free. With Endomondo you can create commitments, but it does not celebrate when the user achieves them, at least I just recognized that, ages ago, I have already created one, and I don’t remember any fireworks, although I completed it 114 times.
Challenges are strange animals in Endomondo since they are actually competitions (e.g. in a distance challenge, runners are ranked according to the distance they run). Although competitions can be very motivating, especially for the racers, I believe that those runners who are on page 1 of a 33000+ pages long list are got somewhat more motivated than those who are on page 13467.
Strava challenges are more of what a user would expect from a challenge: do something well-defined thing (run at least 200k in October). As a result of a challenge, unlike in Endomondo, users are not ranked, but get a trophy if they met the challenge. Trophies are good motivational tools, and as such, help the retention, this is why it is hard to understand, why is it a premium feature to see all trophies in one place.
Runkeeper challenges are similar to Strava challenges but they are easier to complete (e.g. in 2017 October, there are two challenges: complete a run 5k or 10k run), and the finishers will get a discount off of selected items in the Runkeeper store. Interestingly, the user’s challenges are only available on the web.
Clubs in Strava are more for community building: a bunch of people who can discuss many topics, and ranked in many aspects (speed, distance, etc.). The ranking is automatically reset every week.
With Runkeeper we can motivate to run our friends by inviting them to a running group, which has to complete a challenge to a given deadline. Running group can contain up to 25 invited friends of a user. Running groups are not available on the web.
All three apps reward new best times with trophies. This has a great motivational power as long as the runner can improve his best times, but the celebrations will inevitably be getting rarer over time. Strava defers this time somewhat by having route- and segment-based best-times; event if the runner cannot improve his 30k time, he can get faster on certain segments. Besides celebrating yearly and absolute best times, Endomondo also pops up a celebrational screen after user finished a run which a monthly best time (say 15k). However, because there is no or little performance behind certain celebrations (e.g. the longest distance in the given month, every month). Such celebrations can inflate the motivational value of the feature.
Racer concerns
In Endomondo and Strava users can analyze charts and statistics, and make their own deductions and decisions how to train, while Runkeeper helps directly the racers to improve in the context of Asics Academy Challenges. For example, a 5k academy challenge, based on the previous 5k paces, offers a training plan to improve the 5k personal record of the user. But the help in the free version of Runkeeper is limited to the challenges, training plans are paid feature in all three products.
Strava is the only product, which gives some help the racers to prepare for a race. It has a motivational force that they can join a race community and train virtually with others by registering a race event. They can set their goal finishing time (Strava also gives PB forecast), study the profile of the track, discuss the race with others, see the goal finishing time of others, and their workouts.
Common use cases
In all three apps, runners can get feedback during a run by looking at the recording screen of their phones. However, if the phone is strapped to their left arm, as it often is, they have to look at the screen from upside down, which is not really a good experience. Runkeeper recognized it and offers the user to switch to landscape mode. Alas, at least in my case, in landscape mode, the screen was still upside down from my perspective, and I rather switched back to portrait mode.
Similarly, runners can opt to rely on the announcements of the audio coach during the run. Endomondo and Strava work as you’d expect (it is not configurable what the Strava coach should announce though), but Runkeeper announces the requested data in 1.01, 2.01, … kilometers. Strange. Even stranger, if you set that you want to hear the split pace, you will hear something else, but in the last 10 weeks I could not figure it out (but I guess that something else is the current speed when the announcement starts).
Regarding aggregated stats, Endomondo is clearly the winner. It has a dedicated statistics page on the web which offers lots of aggregation controls.
Fitness runner use cases
I have not got reminders to run from Endomondo and Strava in a 7 days period when I did not use them and did not upload any workout to them.
The user can share his workout with all three products on Facebook with significant differences. Strava activity posts are images with a clickable (and quite ugly) link in the description which leads to the activity page in Strava. Endomondo posts are links (can you read the white text on the light background?): after you click on the map the Endomondo workout page is loaded. Runkeeper posts are just images, the user cannot get to the activity web page at Runkeeper, which is quite unfortunate from user experience and business perspective.
Racer use cases
Table: Racer use cases
The free versions of Runkeeper and Endomondo allow runners to set the type of the workout before running; the audio coach announces the instructions before the segments of the workout.
Runners can compare their race performance with the others’ after the race in Strava using the flybys function. I don’t know if this function was designed for this particular use case or not since it seems to be a rather complicated solution of this given problem, but it is not only fun to use, but also useful for race analysis. Participants can identify those runners who had similar performance than they had and analyze their pace over the course, which can give them important cues of what did they good and wrong.
Summary
None of the products address the main concerns of the primary user groups directly and completely, but through a set of features and use cases which give indirect help. For example, if a user wants to lose weight, he can count his calorie intake and the calories burnt during an exercise. Similarly, if a user wants to run a sub 20 minutes 5k, after studying the training techniques, he can do interval training and fartleks, study his training statistics.
This said, Endomondo is the best choice for fitness runners. Its recording screen is polished and extremely easy to use, it offers richer aggregated statistics than the other two product, and celebrates the achievements right away after the user finished a run. Friends can send live messages while a user is running, which both have motivating effects and cheer up the runner.
Strava is for racers and hard-core city runners, who go crazy when the recording does not stop at a traffic light. It has a huge base of professional athletes the user can follow (although the link is not that easy to find on the web), supports the preparation for the upcoming races, forecasts best times, automatically identifies routes, highlights trends on a given route, and allows the user to analyze his performance on the segments of the run.
There is a huge potential in Runkeeper. It offers two features for free that are unavailable in the free versions of the other two products: goal setting and counting the number of steps/minute during a run using just the phone of the user. It also employs lots of techniques to get its users running. There are some rough edges, however, such as the highly inaccurate speed measurements, the often confusing control flow, the strange announcements of the audio coach, the controls that can only be used by pro users (but it is not highlighted, so you tap on them all the time), and the product seems to be undecided which user group should it focus on.
Strategic product analysis
In the upcoming sections, I discuss how the products dealt with the social media aspects, and the strategic considerations. Please note that, if sufficient data is available (e.g. customer behavior metrics) a deeper, more well-founded analysis can be given.
Social networking
The biggest difference among the three products is in the social networking features, ie. the way the products facilitate community building and the characteristics of communities. The social networking features have a significant impact on all metrics discussed below.
Strava users are athletes in their terminology, who want to get better, and the community of athletes is a club. This also means that it is not friendship what organizes the users into communities (i.e. clubs), but their relationship with running. You can join Strava and enter a local club, or club of marathoners, and as a result, you can see the activities of the club members, and the leaderboard of the club. Essentially, when you sign up for Strava, you potentially get connected to all of its (active) users. The connection of two runners is also based on the sport, users don’t have to know each other outside the running world, and their relationship can be asymmetric: you are not friend of another runner, but you can follow him because you want to see his activities on your feed, and the followed user may not know you at all. Also, you don’t like their workouts because liking is more personal, it also indicates (at least partially) liking the other user as well. Instead, Strava users give kudos for the workouts, which is more an appreciation of a remarkable effort; I got more kudos from unknown people in Strava than from my acquaintances. Consequently, Strava users are not separated but form a huge community, and if you don’t have any friends or colleagues in Strava you will find plenty of clubs to join, people to follow, and you will get plenty of kudos. This obviously has a positive impact on all key metrics.
Endomondo users can have friends, and friendship is the main organizing principle, users are connected based on their friendship, they typically have to know each other outside the world of running, and their relationship is symmetrical. They can also get to know other users who join the same challenge (e.g. monthly distance challenge, which has a leaderboard), but that is a secondary social relationship in Endomondo, only the activities of the user’s friends appear on his feed. Users may like each other’s workouts, but it also involves the expression of sympathy to the other user, it is an extremely rare event that users who are not your friend would like your activity. So Endomondo is barely “digitizes” the already existing relationships between the users, but don’t really open up their world or endorse building running communities. Consequently, it can preserve and strengthen existing communities, and therefore it can strengthen the metrics, but since users who don’t already know each other outside Endomondo don’t interact with each other, and therefore the effects of social networking features are limited by the existing social relationships of the runner.
Runkeeper builds on existing social relationships exclusively. We cannot see the participants of challenges, there is no leaderboard. We can become a friend of another user, but since friendship is a symmetric relationship, and assumes that we already know the other user, Runkeeper does not help to extend our social network beyond what we already have outside running. Unsurprisingly, you can only challenge your friends, but you cannot create a challenge for all Runkeeper users and you have no means to build running communities in Runkeeper. Friends may like each other’s activities just like in Endomondo. Consequently, even though Runkeeper has 50 million users when you joint the user base you are also separated from them, and if none of your friends use it, you are basically alone. Therefore, the social features in Runkeeper have the smallest positive impact on the key metrics among the three products.
Runkeeper
Acquisition. In 2016 Runkeeper has been acquired by sportswear giant Asics. This opened a new potential acquisition channel: the product can be advertised to buyers of Asics running shoes and gear and participants of Asics sponsored sports events. Given the limited feature set of the free product, offering a limited time free subscription would increase the acquisition further without threatening revenue.
If you google “run app”, Runkeeper is the only product of the three which appears on the first hit page. Similarly, if you search for “run” in App Store or Play Store, the other two apps are way below on the hit list. This is a clear competitive advantage.
Runkeeper allows sharing workouts on Facebook and Twitter (although I have not set up Facebook connection I got “Activity successfully shared” message after I selected the “Share to Facebook” menu item; yet another quality issue). According to Facebook, more than 500,000 users share their activities on Facebook every month, and given that a Facebook user has more than 350 Facebook friends on average, Runkeeper activity posts potentially reach 175 million people every month, while Runkeeper already has 50 million users.
The signup page is in line with the mission of the product: “get the whole world running”. It’s promise, “helping people get out the door and stick with running”, clearly targets one of the main concerns of fitness runners: keeping up the motivation to run. User testimonials aim to strengthen this promise and build trust and add some further details about the capabilities of the product. However, it also states that Runkeeper is a “community”, which is a mostly unfulfilled promise given that if the user’s friends don’t use the product, the user is left alone.
Activation. Signing up for Runkeeper (using Facebook, Google), installing the app, running the first time, and checking the details of the just finished workout are drop simple. Problems start right away after the workout results are shown: the user can get confused how he can get back to the main screen (with the tabs), and it is hard to wrap one’s head around that the previous screen was the workout dashboard but from the results screen we can get back to “Friends”.
Problems start when the runner wants to check a previous run with the mobile app: it is hidden in the labyrinth of tabs (tap on tab “Me”), charts (scroll down to “Activities”, and tap on it), and lists (scroll down to find the desired run and tap on it). Well, as a matter of fact, this is how I found it the first time and used the app for weeks. But, there is a faster way to get to the feed: tap on tab Friends (yes, you need your Friends to see your feed), and there it is. This might be the first time when a user would feel being lost and anger, which could easily lead to uninstall. The web feed, on the other hand, is the default screen; it is way easier to find a training on the web.
Retention and engagement. Runkeeper employs several tools to get the user hooked on running. Challenges with discount shopping rewards, running academy challenges, runner groups, goals, and notifications to train all serve as triggers.
Fitness runners often need motivation, and there are just a few people in a community who are not only motivated enough to run regularly but also willing to organize that community and give them extra power. Therefore, in theory running groups seem to be excellent motivational tools, but in practice, I feel that they are like Endomondo pep talks: they are fun, but barely anyone uses them. Moreover, the group size is maximized to 25, so in order to cover all users, at least 4% of the runners should create a running group, which seem to be highly unrealistic.
Runners get discounts on selected items in the Runkeeper store if they complete a challenge. This October there are two monthly, easy to complete challenges, which, even in case of single-digit conversion, can result in several hundreds of thousands of purchases in the store. But there are several issues with such challenges: there are too few of them, the bar is too low for more advanced fitness runners and racers, the discount does not apply to all items, and while they are external motivational tools, in practice the reward is more often lost then used (i.e. people don’t buy stuff from the store every month). Besides, the two challenges have slightly more than 100,000 participants together (with obviously non-zero overlap), which is less than 0.2% of the total number of users. Possible reasons: the number of active users on Runkeeper is way less than 50 million, users cannot see the challenges on the web page (they are also missing from the activity feed on the web), users are not interested in challenges in general, users are not interested in easy to complete challenges. In summary, regular Runkeeper challenges have a negligible impact on retention.
Goals, and especially running academy challenges are stronger, internal (and external, see triggers) motivational tools, which are missing from the free versions of the other two products. There are more than 330,000 participants of the 5K running academy challenge (this fact suggests that regular challenges are simply not interesting enough for the users). Running academy challenges cleverly pull fitness runners towards the racer identity, and ultimately, subscription. They also build trust towards the product, and since an academy challenge consists of several trainings that can be done several times, helps retention.
Note, that the pre-requisites of high-engagement are useful and fun features and services. Usability issues, inaccurate data, mobile-only features (e.g. running groups and challenges) weaken engagement. Also, can the user spend much time to analyze his runs, get training tips, follow professional runners and study their workouts, prepare for a race or analyze a past race (including the performance of other participants)? Unfortunately, the answer is no. Finally, since it is not always obvious which functions are available for free in the mobile app, the user taps on paid features way too often, which lead to pages offering subscription. This can annoy most of the new users, and lead to lower retention.
Monetization. Runkeeper challenges route users to the Asics store, but the fact itself that Runkeeper is part of Asics increases the chance that a user will buy Asics products.
Runkeeper offers advertising partnership to companies, thus they manage to monetize their user base. They offer sponsorship-like advertising models like workout rewards and custom challenges, which are also beneficial for the users. I deem this a very good model, but that I have not seen any ads or sponsored rewards I suspect that it is not a very successful one.
Runkeeper also pushes users to subscribe. The paid services might be useful, but the way and number of occasions a user bumps into the subscription offers make him more frustrated and angry than willing to pay for the pro version. Also, the free version of the product seems to focus on fitness runners primarily, who are less probably spend much money on running compared to a racer, and still, Runkeeper tries to sell such features for them which are more interesting for racers. This contradiction must hurt monetization. (An illustration of this issue is the image attached to the subscription offer, which displays a runner who just did a short slow jog.)
But if we assume that the proportion of subscribers among all users is around the same as the as the proportion of racers among all runners then we get that 1% of all users, around 500,000 runners pay the subscription fee every month. This is twice as much as in case of Endomondo (estimated) and in case of Strava (data from 2015).
Runkeeper doesn’t seem to collect much money from application developers, at least I did not find an app store or list of applications that are built on it from my Runkeeper page. But after googled “pro athlete runkeeper” the hit list contained a link to an app page under runkeeper.com, and after I cut the app specific part from the URL I landed on the app store page.
In summary, the small number of challenge participants, and the lack of advertisers hint that the active user base (those, who run with Runkeeper at least a few times a month) must be even an order of magnitude smaller than the number of registered users. Subscription can be the main source of Runkeeper (as opposed to Asics) revenue; based on the number of registered users I estimated the number of subscribers to 500,000, but given the discrepancies in user focus and what they try to sell, the number of subscribers can be much smaller.
Referral. Runkeeper allows sharing runs on Facebook and Twitter, but the number of running enthusiasts who are sharing their Runkeeper workouts is the key factor. Besides, the share button is hard to find on the web, it is hidden in a menu of the training, and the Facebook post is just an image, the friends of the user cannot get to the activity page.
Since trust and authenticity are extremely important in this context, and besides one’s friends, pro athletes can have the necessary convincing power, I googled “athlete using runkeeper” without much result.
Churn. The other two products offer more features and better user experience than Runkeeper. However, Runkeeper is the only app which offers a free training plan, which can be appealing those fitness runners who want to train more conscientiously.
Besides, the monthly subscription fee is just 4.99 USD (in October 2017), while it is 5.99 USD in case of Endomondo and 7.99 USD in case of Strava. Is this price different appealing enough for a racer to migrate from Strava to Runkeeper? I doubt that; running is their life/hobby after all, and money plays relatively little role in this context.
Endomondo
Acquisition. The user acquisition power is clearly the weakest of the three products. Although Endomondo has been acquired by sports and casual apparel manufacturer Under Armour two years ago, this did not open a new acquisition channel, since according to my research, Under Armour has not sponsored a major running sports event in 2017, and the number of sponsored runners seem to be zero. Besides, if you google “run app”, Endomondo does not appear on the first 5 hit pages, and when you search for “run” in App Store or Play Store, you can not scroll down enough to find Endomondo. On the plus side, Endomondo is well-ranked on the lists of posts about top sports trackers, and these posts appear on page one of “run app” Google hit list.
User acquisition relies on referrals significantly. According to Facebook, more than 1 million users share their activities on Facebook every month, and given that a Facebook user has more than 350 Facebook friends on average, Endomondo activity posts potentially reach 350 million people every month, the same number as in case of Strava, and twice as many as Runkeeper. Endomondo has 25 million users, half as many as Runkeeper (as of 2015 March, Strava had 1.2 million active users).
The signup page targets the fitness runners: it promises fun in running and help to stay motivated. There are no user testimonials, and fitness is not identified as a goal.
Activation. It is super-simple to sign up for Endomondo (using Facebook, Google), install the app, run the first time, and check training data on mobile or the web. The user experience in this respect is polished almost to perfection.
After the first run, users can easily get familiar with the web page and find what they are looking for. The mobile app uses navigation drawer, however, which may have a negative impact on the engagement.
Retention and engagement. Challenges and celebrations of new achievements (absolute/yearly/monthly best times or distances) after the workouts are the strongest motivational tools in Endomondo. Challenges, however, have their weaknesses: they are sub-optimal internal and external motivational tools. Concretely, challenges are rather races, and farther a runner is from the top positions, the less may feel to be encouraged, and there are little or any challenges that offer discounts or any tangible awards. So maybe most of the racers can get motivation from challenges, but their number is dwarfed by fitness runners, and since they regularly train for a race, they get to rely on less external motivation tools like challenges.
The biggest challenge in Endomondo has roughly 500,000 participants, which is an order of magnitude more than in case of Runkeeper, which may suggest that the number of active users in Endomondo is roughly 10 times bigger than in case of Runkeeper. However, out of the half million participants at least 200,000 have not run this year with Endomondo, and on average only roughly 17 hours were spent on running in 10 months by the participants.
New achievement celebration screens can boost the pride the runner and encourage him to run again, but not all achievement have the same value for the user, and the value monthly celebrations can inflate over time.
The more workouts a user has per week, the more time he will spend on analysis. Around position 20,000 of the most popular challenge the users run about 3 times a week, and they probably spend around 5 minutes after each training checking data. The abundance of various statistics, charts, and filtering options primarily get the racers hooked on the product, they could easily spend one hour with analyzing data every week.
In summary, challenges and analytics-focusing engagement optimization resonate more with the needs of the racers, and the bulk of the users, those, who are targeted by the mission of the product, the fitness runners, may need other functions to increase their time spent on running and engagement with the product.
Monetization. The Under Armour Store is available with one click from Endomondo website, but given that there are no use cases that direct users to the store, I assume that not many customers arrive at the store from Endomondo.
There are tens of thousands users who train at least 3 times a week and spend at least 10–15 minutes on using Endomondo outside workouts. This is not particularly much in itself, but the fact that the user base is relatively well off, health conscientious, and do sports regularly, makes their time spent with the product more valuable for the advertisers. Unlike Runkeeper, Endomondo serves ads with Google AdSense. The ads don’t seem to particularly utilize the specificity of the users.
In the US, roughly one-fifth of desktop/laptop users and one-tenth of mobile users have an adblocker, and their relative proportion increases year by year potentially threatening the ad-based revenue stream. In order to mitigate this threat Endomondo could build its advertisement partnership program and move more towards challenge sponsorship oriented advertisement, but as we have seen at Runkeeper, being part of a giant sportswear manufacturer may not particularly make the product appealing to advertisers even if it has a considerable size active user base.
Roughly every 6th user has a subscription among the top 500 participants of the most popular challenge, and the proportion of pro users falls significantly behind them. Endomondo is already rich of features, and the premium services are targeting racers who are roughly 1% of the runners worldwide, so it makes sense to estimate the number of premium users as about 1% of the total user base, which gives 250,000 subscriptions.
Endomondo reached profitability first in 2013.
In summary, the potential for the acquisition by Under Armour has not exploited yet, but Endomondo is already profitable thanks to advertisements and subscriptions.
Referral. Endomondo has hundreds of thousands of active users, and there are roughly 1 million users sharing their activities on Facebook every month, this is twice as much as in case of Runkeeper. I googled “athlete using endomondo” without much result, which means that Endomondo relies on the recommendation of non-professional runners.
Churn. Endomondo offers a wide range of features which can be appealing to many fitness runners already using another product. Given the number of active Endomondo runners, Endomondo can attract users of other products with fewer features. Unfortunately, even if Endomondo is superior to another product regarding the feature set (or any other aspect), workouts shared on Facebook don’t reveal this fact.
Racers, especially those, who are already subscribed to another product may find the pricing of Endomondo subscription appealing, but given that not all of their needs are addressed by the product, it is unlikely that they’d switch to save money.
Friends on Endomondo make a user more reluctant to move to other product: he would miss their likes and comments.
There is a real threat, however, that some users leave the product. Roughly two years ago, due to the changes of the calendar and workout display, angry comments flooded the Endomondo blog and Facebook threatening to switch to another product. Although the company reacted relatively quickly, this event highlighted, how fragile the customer experience, and eventually, the loyalty of users can be. Every similar even is an opportunity for the competition to attract users, especially if they offer better user experience and more (useful) features.
Strava
Acquisition. Strava is the only independent product of the three. As we have seen in the case of the other two products, acquisition and retention are not necessarily benefit from the tailwind of the mother companies.
Googling “run app” returns Strava on page two, but is there anyone who checks the results on the second page? On the other hand, Strava usually is in the top 3 of run tracker listings which appear on page one of the Google hit list. In the App Store and Play Store Strava is also way from the top hits when you search for “run”, but it can be found after scrolling down a little. Again, there are not many people who are willing to scroll down that much unless they are explicitly looking for something.
Similarly to Endomondo, Strava has 1 million monthly users who share their activities, and by calculating with 350 Facebook friends on average, this means that Strava activity posts reach 350 million people every month. As of March 2015, Strava had 1.2 million active users and 200,000 premium users. The number of active users is dwarfed by the 25 million Endomondo, and 50 million Runkeeper users, but those figures contain non-active users, and as we have seen from the number of Endomondo and Runkeeper challenge participants, their number of active users cannot be significantly higher than 1 million (it reaches this number at all).
The strongest acquisition power of Strava stems from the large numbers of professional athletes who are using the product (see also the 2016 stats); they are very strong assets for the brand, borrow authenticity to the product, and build trust among the existing and future users. (Strangely, however, the link to the professional athlete Strava page is almost hidden on the bottom of the user’s profile page beside “help”, “about” and similar links.)
The Strava signup page promises connecting athletes, in line with their mission. Unlike the other two products, it does not promise help in staying motivated and getting fit, but states that “if you’re active, Strava was made for you”, and it offers users to “get better by analysis”. These focus extremely well on the racers, maybe so much, that they may also frighten off the fitness runners. Many people start running to preserve their health, and become racers after they get hooked on running, however Strava don’t try to acquire them when they just start running, which means that racer acquisition relies on the churn of other products, but if someone got used to a product, he will reluctantly abandon it for another one. Given the significantly higher number of fitness runners it surely hurts the acquisition power of the product, but given that racers are already active, it may benefit the retention and engagement. (Besides, Strava has to invest less into the development of motivational features.)
Activation. Just like in the case of the other two product, signing up for Strava (with Facebook, Google, email) and installing the app are effortless. But while Endomondo and Runkeeper apps start with the recording screen, and all the user need to do before a run is to push the record button, the Strava app displays the user’s activity feed. Well, “Strava is the social network for athletes” after all, but this design makes activation a little bumpy, as the user has to find and switch to the recording tab. Again, this may not hurt the activation among racers but can lead to a large number of fitness runners launching Runkeeper or Endomondo instead.
After the first run, the app automatically switches back to the feed and the user has to tap on the training’s entry to see its details (just like on the web). The structure of the web and app pages is surprisingly deep (it was, after the flattened design of Endomondo).
Compared to Endomondo and Runkeeper, Strava starts with a significantly positive experience during the first run for city runners: the auto-pause (if activated) kicks in as soon as the user stops.
Retention and engagement. As we could see right away on the sign-up page, Strava expects users to be already active, therefore it puts less emphasis on motivating them externally. There are several motivation tools that apply Strava though: runners can join running clubs, complete challenges, give kudos to each other for a training, get rewards for improving their performance of segments and routes. This October challenge rewards range from getting digital finisher badge to participating in a draw for a $250 gift from Reebok (who understandably does not sponsor a challenge on the run trackers of Asics and Under Armour). Monthly running challenges usually have 150,000 participants, half-marathon running challenges have roughly 100,000 participants (note that the biggest Endomondo challenge has 500,000 participants, but that is a yearly challenge), and almost all of them have at least one run in that month. These facts also indicate that Strava users are more active, but also indicate that just a small proportion of all active users actually join a challenge.
Strava provides lots of data to analyze, but unlike Endomondo, it does not offer aggregated statistics. Instead, users can submerge into the lists and charts of route- and segment-based pace/elevation/heart rate data, and analyze their performance over time and during a given run using the large number of segments (which they can create, and make public so that others can check their performance in that segment). So segments not only motivate runners by ranking them but also get users hooked on the product itself both on web or mobile.
Runners can compare their performance with other runners, and see who ran where and when they were out for training using the flybys function. During a regular solo training flybys are more like a time-wasting game (which is good for engagement), but it can be very handy in evaluating one’s race performance, since the runner can easily identify those Strava users who ran the same race with around the same pace, and learn from them by analyzing their data. In general, flybys, running clubs, and following professional athletes can serve as jumping boards towards analyzing each other’s runs, which has a positive impact on retention and engagement outside the actual workouts.
It is strange that users cannot set who did the run with. It would fit the mission (social network for athletes), it would allow to build the community and find runners with similar performance and analyze their workouts, and it would also reduce churn (both runners have to use the same product). Since racers always prepare for the next race, the race event function is also a useful tool for them, and for increasing retention. They can spend hours with analyzing the track and elevation profile of the race, follow the workouts of other participants, and analyze their runs.
In summary, Strava functions are clearly designed for racers, and although Strava assumes its users to be motivated, it employs several motivational tools for active runners who like to compete and would like to improve. Runners can spend several hours per week with analyzing their own runs, they can easily find those users who perform similarly as they do and analyze their runs, and they can also see the data from the runs of their favorite pro athletes.
Monetization. According to Strava CEO James Quarles, “Trying to do banner ads and things like that, it just doesn’t feel like it dovetails with what people are opening the app for.” So, even though Strava users have plenty of opportunities to spend time with the app and the web, and they seem to do so, Strava doesn’t monetize their time with ads.
However, Strava monetizes the data with its metro business: they sell aggregated and anonymized commute data to “departments of transportation so they can better plan pedestrian bicycle routes in cities”.
Strava primarily makes money with subscriptions. Their monthly subscription fee, 7.99 USD, is the highest of the three products (Runkeeper: 4.99 USD, Endomondo: 5.99 USD), but despite this, in 2015 March they already had about 200,000 subscribers (the estimated numbers of subscribers in Endomondo and Runkeeper are 250,000 and 500,000, respectively).
Strava is surrounded by a wide ecosystem of apps, however, as of today, they give API access to developers free of charge.
Strava has not reached profitability yet.
Referral. The biggest competitive advantage of Strava is the user base of professional athletes, who refer to Strava in their blogs and vlogs. Friends’ posts on Facebook and Twitter also build trust in the product and strengthens the Strava brand as a product for enthusiast and professional athletes.
Churn. Strava focuses on racers, therefore it is highly unlikely that it will attract fitness runners from other products. Racers, however, may look around when they decide to subscribe to the product they are already using and consider switching to Strava. The analysis of the pro feature set is out of the scope of this post, and it is hard to judge if the customized training plans and the advanced analytics are really worth the monthly extra 3 and 2 USD per month compared to Runkeeper and Endomondo subscriptions in addition to the inconvenience of learning a new product.
In theory, free Strava users may switch to the other two products when they decide to subscribe due to the price difference; in practice, however, the price difference is almost negligible.
Summary
The product strategy of Runkeeper and Endomondo must be aligned with the business strategy of Asics and Under Armour, respectively. Asics has already sponsored the New York City marathon, and sponsors marathon and track runners, while Under Armour seem to focus more on fitness and other sports such as basketball.
Runkeeper has great acquisition power: it is in the top places of search lists, Asics is well-known and honored in runners’ world, its signup page delivers well-phrased messages to fitness runners. The activation is bumpier due to UX glitches and quality issues (such as inaccurate data and audio coach announcements). Runkeeper offers the least charts and data of the three, but it also employs the most motivational tools of the three: the first hurts the engagement among racers, the latter helps the retention among fitness runners. While in absolute numbers Runkeeper could have the most subscribers, its user base does not seem to be active, which could impact the ad-based revenue of the product. The inactivity of the user base also impacts the referrals negatively, and there is also a danger that certain runners decide to switch to another, a better quality product with more features. Also, Runkeeper does not help to build running communities, social relationships are coming from the outside world; consequently the social networking features of Runkeeper has a very limited positive impact on the key metrics.
Endomondo, although it has 25 million users and despite that there is a huge unexploited potential in the cooperation with Under Armour, has the weakest acquisition power compared with the other two products in both user groups: it is not among the top search hits, it is not used by professional athletes, but it relies on its relatively active fitness runner users, who already know each other outside Endomondo. Once a runner finds Endomondo, he will almost certainly register, and get a great experience during and after his first workout. Its analytical capabilities engage racers more, while it could provide motivational triggers for fitness runners. User engagement is monetized by advertisements and racers are offered a good value for price subscription, however, there is no function of the product which would directly relate to the running events (note: there was an event calendar years ago). Again, there are no signs of cooperation with Under Armour. The fitness runner user base seems to be active, which helps the referrals, and they have little reason to leave the product, but the racers may consider to switch to Strava even though their subscription fee is higher (but then they’d have to leave their Endomondo friends).
Strava seems to give up the fitness runners and focus on racers exclusively. Although the other two apps are backed by sportswear manufacturer giants, its competitive advantage is the big professional user base which helps user acquisition among racers. The activation is relatively complicated compared to the other two products. Users can spend considerable time on analyzing data which is appealing for racers, while the other two products implement better motivational tools for fitness runners. Strava is not profitable yet, they decided to be ad-free and rely on subscription, which is the most expensive of the three and selling aggregated, anonymized data. Its social features and professional athletes both help to achieve a high number of referrals and low churn. Strava facilitates building running communities, even if you don’t know any Strava users, you will be able to build your running social network inside Strava, which has a positive impact on the user behavior metrics.
Proposals
Finally, in the last sections, I propose some improvements in the products based on the findings of the user-oriented and strategic analysis.
Runkeeper
Runkeeper has several strategic advantages ahead of other products: it is at the top of any search list related to run tracking, it has a huge user base, and it is part of Asics, which is a well-known brand in the world of fitness runners and racers.
We don’t know exactly what percentage of the 50 million Runkeeper users are active, but given the quality issues, the relatively low number of challenge participants (on November 3, the November Runkeeper 10k challenge has 15,108 participants while the similar Strava challenge has 117,362, almost 8 times more), and the relatively low number of Runkeeper Facebook users (about 500,000) I estimate it well below the number of Strava active users (1.2 million). This said, the sheer number of registered users is a great value in itself, even if most of them switched to another product (which is very likely) or stopped running (which is even more likely) because after well-designed improvements many of them can be enticed back to Runkeeper. Moreover, the Runkeeper brand can be strengthened among racers if runners sponsored by Asics start using Runkeeper publicly, which can increase user re-acquisition and re-activation among fitness runners. Built on the re-activated user base, the Asics brand, and improved pro functionality, the number of subscriptions can also significantly increased.
Runkeeper, however, suffers from quality issues which, I deem, endangers the future success of the product. Therefore, first of all, quality issues shall be fixed, including data inaccuracy, the strange behavior of the audio coach, and UX glitches. Besides, the today bare fitness feed must be improved so that each running workout entry contains at least the map, the distance, the duration, the average pace and the elevation. The feed must be super fancy from the perspective of a fitness runner!
Introduce a virtual currency to increase retention and referral, and reduce churn. A user would get points after he completed a challenge, finished a race, or shared a workout on Twitter or Facebook, then he could spend the collected points in the Runkeeper/Asics web-shop: buy items, exchange it to discounts, buy pro subscription or participation to an Asics sponsored race. Besides, in order to keep users active, the collected virtual should inflate over time, but should not melt below a minimum value in order to reduce churn. This would make Runkeeper more visible of social media, make harder to switch to another product, but would also motivate users to run.
Open up the Runkeeper community and allow find virtual running buddies in order to activate runners, and increase retention and engagement. Introduce a Runkeeper community feed which would show the workouts of certain users: users who are in the same area, who have the same performance as the logged in user, or have outstanding performance, or run on an interesting track. By using machine learning, the user’s behavior could influence the future contents of the track, if he tends to click on workouts with specific properties, it will be worth to show more of those workouts. The feed should be real-time in that sense that it should be updated automatically on the screen to suggest that Runkeeper has an active community, which never stops. Also, the users could use the community feed to find running buddies all over the world. Running buddies are virtual friends with whom the user may never meet in person, but revel in each other’s accomplishments, encourage together, give advice, give kudos, challenge each other. So running buddies are not friends imported from the user’s private life, but people the user met in Runkeeper, they form a private club for the user. Just like Runkeeper friends, the user could challenge them. Well, as a matter of fact, I’d also rename friends to running buddy in the product.
Other “random” recommendations:
- Mark the pro features or remove them completely from the screen (Improves UX and activation; Reduces churn)
- Extend the “Upgrade” screens (which offer subscription and give a textual description of the pro function the user tapped on) with an option that would allow the user to see actual sample screens of the corresponding pro function(s) (e..g. Lifetime stats). (Improves monetization)
- Make Prescribed Workouts available in the free version to strengthen the commitment of fitness runners and therefore strengthen retention, reduce the churn of racers and increase engagement. The feature should be available on the web also. (Improves retention and engagement; Reduces churn among racers)
- Recommend running events to the users based on their location and performance. (Improves retention, engagement, and monetization)
Endomondo
Endomondo has a relatively active user base, and features that are appreciated by racers as well, but Under Armour seems to focus less on running than the other two companies, which is a strategic disadvantage if Endomondo wants to focus on racers. On the other hand, Endomondo has a strategic advantage in targeting fitness runners who’d want fun and encouragement.
Provide a platform for running game developers to improve retention and engagement among fitness runners, and improve monetization by opening a new business line. Running games would receive the current location and speed, and the workout route of the user, and based on this information they could access to the speakers and could add and update a game-specific an overlay to the map in the mobile app; they could also have a leaderboard on the web. An example game: the user needs to visit certain places on the map and collect points based on the time he needed to get to those places. These games could be free or paid (in-app purchase); Endomondo would deduce certain percentage of the price of the app, developers should pay a yearly fee for the access to the Endomondo gaming platform. Endomondo would be able to improve monetization and utilize the creative forces of external developers; the external developer would gain access to 25 million users and an active user base; games would help users to stay motivated, improve retention and engagement.
Introduce monthly and yearly improvement challenges sponsored by Under Armour to improve retention and engagement. A participant of an improvement challenge would have to improve a certain personal best time or distance (i.e. 10k PB in the previous month, all-time longest distance, this year’s best 5K) by a certain percentage (e.g. 7%). There should be plenty of such challenges in every month in order to cover that most of the user base. Each challenge should have a reward which could be exchanged for items or discounts in the Under Armour store including pro subscription. Such challenges, due to the tangible reward and the relative goal (i.e. it is not the fastest wins, but the one who gets better the most), would be more appealing for fitness runners than the common challenges. Besides, it would also make training plans, that are available for subscription, more appealing and could turn fitness runners into racers.
Celebrate improvement success stories to encourage fitness runners and improve retention and engagement. Endomondo should find those runners who were inactive or in the bottom 10% of the active runners and improved the most in a certain time period (quarter, year) and with their approval celebrate their improvements in the Endomondo community in a success stories feed, on Facebook and in motivational newsletter sent to inactive or bottom 30% users. Such stories would be auto-generated, but users could add some personal notes to it when they approve the publication of their story.
Other “random” recommendations:
- Get rid of the gray text on gray/white background color schema. (Improves UX)
- Buyers of Under Armour products should get three months of pro subscription. (Improves acquisition)
- Re-introduce event calendar. Users should be able to create and join events. (Improves engagement)
- Mesure shoe mileage, and warn the user when it exceeds a threshold. (Improves retention)
- Warn if the running mileage increases too fast to prevent injury. (Improves retention)
- Show calories burnt as amounts of various foods eaten. (Improves engagement)
- Make a strategic partnership with a food magazine or blog (such as Cooking Light): if a user requested, they would post nutrition advice to his feed. (Improves engagement)
Strava
Strava has the necessary potential to become the go-to product for a racer. Also, since Strava is not part of a big shoe and sportswear corporation, it can position itself as an independent community of passionate athletes who are willing to invest time and money into running.
On-time purchase of training plans to improve acquisition and monetization among racers: On signup, the user would have the option to select his upcoming race, and depending on the race, offer a free 10k training plan or a half marathon/marathon training plan for the price of the one-month subscription (7.99 USD). Also, if an existing user registers a race, he should be given the same offer. The purchase of a training plan should not unlock all features which are bound to subscription, but only those which are necessary to execute the purchased plan.
This improvement has several benefits. First of all, it helps to convince the users of another product who prepare for a race (which might be their first marathon) that it is worth to switch to Strava, since it directly addresses their problem (prepare for the race), without longer-term engagement (subscription) for a reasonable price. It can also give a teaser of the capabilities of Strava so that such one-time customers could be converted to subscribers. Besides, this offer also addresses those many Strava users whom today don’t subscribe to the premium services, but would happily pay for a help in marathon preparation.
More challenges coupled with training plan offers and challenge sponsorship bidding to improve monetization: Although Strava target racers, the monthly half-marathon challenge is the hardest one. There should be various monthly, quarterly, and yearly distance and speed challenges which are harder to complete. For example, run six marathons this year, or run 1200k during the summer, or run 5k in 20 minutes. There will be a significant number of users who have similar ambitious goals, and by accepting these challenges he would openly commit to achieving them. These challenges should be coupled with training plan offers (like above) which would help the user to complete the challenge.
Besides, each challenge could have a local/global sponsor, who would offer a reward for those who complete the challenge. Registered sponsors would have the opportunity to bid for the sponsorship of the upcoming rewards (e.g. using generalized second-price auction). Local sponsors could be small companies, local sportswear shops who would not advertise otherwise (especially not globally) but would be willing to let those people who live nearby to know about them. Such local sponsors could offer a reward for those users who complete a given challenge in a given area (e.g. in Budapest). This would not only open a new revenue stream for Strava, but it would be beneficial for both users and independent advertisers who could get tangible rewards for their achievements and would get to their potential customers for a reasonable price, respectively.
Facilitate buying running shoes and equipment through Strava retail partnership program. Today when a shoe expires, the user has to go to the nearby store or the web-shop which takes several steps (type in the address of the web-shop, register, search for the product, …). Similarly, if a runner looks for a sports watch he can check what is used by the professional athletes, but then he has to google it, read the specs, find a shop, etc. When a user clicks on a running gear on the web and in the mobile app, the product specs page would be brought up. As part of the retail partnership program, local stores and web-shops then could register with their product prices at Strava, which would display them based on the location of the user, in the order of product prices.
Users would save time with clickable product links because they would not have to google for the product specs; besides, they would find the store with the lowest price. Local stores and web-shops could increase sales via the retail partnership program. Finally, Strava could open a new revenue stream by charging for the partnership program.
(According to the October distance challenge the 152,000 participants ran 102k on average, which indicates that, on average, their running shoes expire in every 12 months. Calculating with 100 USD shoes and 1% revenue realized at Strava from all purchases we get 152,000 USD revenue per year from just facilitating of shoe selling.)
Other “random” recommendations:
- Improve user experience: Of course, UX is a continuous process, but it should be obvious what is clickable, the training calendar is near useless, just to mention two major issues.
- Recommend route for race preparation based on similarity to the race profile for those who travel to the race (segment X in your town is like segment Y on the race).
- Recommend pro to follow. (Improves engagement)
- Recommend club to join. (Improves engagement)
- Recommend training plans for in-app purchase based on previous performance. For example: “Your best half marathon time is 1:40, but if you complete this training, you can improve your race finish time to 1:33.” (Improves monetization)
- Premium API access to all user data (unlimited, even real-time data). Note that location data must not be for sale, and data should be anonymized. (Improves monetization)
- Become the platform for running coaches and their coachees. Coaches could check the workouts of their runners and submit them (updated) training plans, and exchange messages with them. (Improves monetization and engagement)
Liked this post?
If you enjoyed this post, please help out your friends with a *clap* or a share.
You may also want to visit the Product School home page or Youtube channel.
