10 Jan 2019
Exist annual survey 2018
As part of our values of working transparently, we send an annual survey to Exist users to get their feedback on how we're doing and what we should work on next.
In practice, each has been a bit more than a year apart, as there's always something in progress we want to get done first. This year it was the Garmin integration.
So without further ado, let's go through the survey questions and your responses. This year we received 244 responses. If you're not interested in individual questions, feel free to jump to the bottom where I'll summarise and talk about how we'll be acting on your feedback.
This year 41% of respondents use the Android app, and 53% use the iOS app. This is a much closer ratio than the previous year, where 24% used Android and 70% used iOS. As the person doing our Android development this is nice to see.
The main reason respondents use Exist is because it "offers interesting or useful insights into their behaviour", with 54% picking this option. I'm glad this was number one, that's what we're aiming for! "Keeping data in one place" was next at 23%, then less popular answers such as "supporting Hello Code" and "managing an illness or medical symptoms".
The biggest problem respondents have with Exist is "too much effort to track mood and tags" at 25%. I'm not surprised here, as we've had people mention the effort of manual entry ever since we added the option—this is why most data in Exist is synced automatically. We're working on some ideas about changing the mobile review screens to make this process easier, however. The next popular answers were "too many obvious correlations" at 16%, and "not enough information" and "not enough integrations I use", tied at 12%. Free-form answers to this question mostly tied into "obvious correlations" and "too much effort" choices, with a few "lack of manual tracking" answers.
The most popular features added in the last year were historical browsing and the new mobile summary screen, both at 44% (this question allowed multiple answers). I'm glad these were well received! We had a few vocal users complaining about the mobile redesign, and complaining about taking time to work on the web app, so it's nice that they don't represent the majority of users (and let me tell you a secret—the most vocal, least polite users rarely do). Close behind at 39% was renaming custom tags, then a smattering of iOS-related choices like mobile graphs and new Apple Health attributes. Apple Health is probably our most popular integration these days, so this makes sense.
Mood tracking and custom tags were rated the most useful features in Exist overall, both at 76% (multiple answers allowed). Folks have a love/hate relationship with their daily review—they find it the most valuable, but also the hardest to do. You've got to put in the effort to get the reward I guess! Next were the web dashboard and correlations at 50%.
Most users found insights at least somewhat useful. When asked how useful they found insights on a scale of 1–5, 41% of users picked 4 or 5, and 41% picked 3/5.
The most popular potential addition was "habit tracking features" at 72% (multiple answers allowed). It makes sense that users want to expand their use of tags to replace habit tracker apps entirely, but I'm not sure we'll have time to add this in the next twelve months. Next was "automatic venue tracking" at 50%, then a lot of other choices ended up around the 30–40% mark. I'm not surprised about the venue tracking, as Moves used to do this and then was killed off by Facebook acquisition, and now Gyroscope offers something similar I think. Still, this level of location tracking creeps us out, to be honest, and I don't think we want to take on the risk of tracking and storing this data on users.
The two most popular categories for new integrations were "more data from existing integrations" and "productivity", both at 32% and clear winners ahead of a large spread of less popular responses. Great! The former is a lot quicker than adding new integrations, so thanks for considering my workload. I can also see why productivity is popular, as we have a lot of productivity attributes only served by one integration (e.g. tasks completed from Todoist, emails from Gmail). We can work to expand the options here.
The most popular focus for our work on Exist for the next year was to improve correlations, at 30%. This included examples like adding next-day correlations, rating correlations positively or negatively in mobile apps, and anything else that will make them more useful. Improving correlations has been a common theme throughout Exist's existence, and something we're always trying different ideas to achieve. Still more to do! Multiple values per day (e.g. multiple mood ratings) was next at 20%, then manual tracking (or allowing tags to have numeric values) at 17%. We've noticed a few people asking for (or often, being furious about the lack of) manual tracking, and this option is popular on the roadmap, so it's surprising that it wasn't the first or second option. When it came down to picking just one, most folks thought it wasn't top priority.
88% of respondents have custom tracking turned on, up from 75% the previous year. Most of those users add tags just once at the end of the day, at 69%.
Mood tracking is used by 96% of respondents, up from 93% last year. 7% do not use the mobile apps to rate their day, and 43% sometimes catch up by rating multiple days at once.
About 4% of respondents have written a custom integration or script to send data to Exist. Which is really cool! Shortcuts on iOS might drive this number higher in future.
Around 7% have used a community-made integration or shortcut from someone else. And 37% of respondents have checked out our API docs, which is surprisingly high.
We then asked respondents how they felt about our values, which were things like "we have a transparent roadmap" and "we don't have investors to tell us what to do with your data". And all of our values had a majority positive response, thank you! I'm glad these values are appreciated, because they come with tradeoffs for us, like not being rich 😉 The largest negative response was to "We don't use analytics at all in our mobile apps", with 4% of respondents not liking this. Fair enough, it does mean it's harder to debug mobile issues.
Finally, respondents wanted to read more on "Quantified Self and apps for tracking your life", and "ways to use Exist and interviews with other users", both at 67%. The Exist blog was the most popular content resource, with the least number of users listening to our podcast.
How we'll act on those results
We're always trying to balance what the survey tells us, what's popular on the roadmap, and our own internal plans for improvements. Combining all three gives us the following to focus on for the year.
We'll add the ability to rate correlations to the mobile apps, so you can hide the most obvious or confusing ones (which is already a feature in the web app). We'll also finish next-day correlations so you can see if things you did yesterday affect today's behaviour. We'll find new ways to make correlations useful too, like in the new "What affects your mood?" graph on the Trends page, and giving you the ability to pick any two attributes and see the result.
We'll revamp the mobile interface for rating your day and adding tags, making it faster overall and easier to do things like add one tag without going through all the current steps.
Integrations and attributes
We'll work out where we can get more data from existing integrations, especially things that can be filled by multiple services, and add those attributes. We'll add some more Apple Health attributes as part of this. Looking at the most popular productivity suggestions on the roadmap, we'll endeavour to integrate one or two of those as well.
We'll keep working on the planned expansion of the Trends page into multiple pages for different groups, giving you new ways to see the most interesting breakdowns of your data. We'll try and get these into the mobile apps too, as not enough folks visit the web app regularly. Given the regular positive response to the yearly review email, we'll also look at setting up something similar to send out quarterly.
And that's it! Thank you to everyone who spent the time filling out our (slightly too long) survey this year, and a special thank you to everyone who left us kind feedback. Thanks for all your support.