15 Aug 2014
Josh
Existing month to month, July 2014
This is our second monthly report on Hello Code's progress as a business and our journey building Exist. You can catch up by reading the first here.
Let's get into July's happenings.
Development
July saw a lot less time spent on bug fixing and some more progress on features. Much better. In addition to adding some more content pages like the FAQs, and some small tweaks like client-side validation on the sign up form, the two main features added were the API and the change to the mood email.
We've been debating the relative merits of exposing our API ever since we started building Exist. I knew we'd build a platform that other developers could build on eventually, I just didn't know when that would happen. In July we decided that now is as good a time as any. The API is read-only, and available to our paid users, but it's a fairly comprehensive means of grabbing your Exist numbers and insights. People have been asking us about an API right from the start, so I was hoping that its release would inspire plenty of visualisations and widgets being built. I put a lot of effort into the API docs, because there's no point releasing an API if people don't understand how to use it.
I can count the number of people who've accessed the API on one hand. I don't even need to use my thumb.
Long-term I'm still positive it was worth the effort, and we'll see more people using it as our user base increases.
Changing the mood email was much more controversial. We use Mailgun to process incoming emails and send them to an Exist webhook, ready for parsing. Before the change, a user would get an email asking them to reply with a number out of 5, and a note, each on their own line. We then relied on Mailgun to strip out the signature and the original message, leaving just the juicy content, which we would parse to find a digit and some text. In practice this was problematic. Some email clients, rather than using line breaks or paragraph tags, would wrap each line in a div. What! Why? Some clients would be able to smuggle their signature through Mailgun, meaning it got tacked on to the note from the user. Various clients had various issues, and a lot of last month's bug fixes involved chasing these down. So we decided the easiest fix was to stop parsing email. Regular expressions are for suckers.
The old and new mood emails
The new approach asks the user to tap a link for their rating, opening a new browser tab where they can then add a note if required. Much less error-prone, but more effort for users. We had some disappointed responses from people who missed the simplicity of just replying to an email, which I totally sympathise with. In the end, users can still reply to the email if they like (I must confess I still do this), we just don't publicise that it's still possible. We're keeping the old-school happy and giving everyone else a nicer (and prettier!) experience.
Marketing
Belle spent much of the month finding paid work, and picked up a few new freelance clients. This is the unfortunate downside of Exist not paying us much yet, and means that her marketing output was less than usual.
In July we kept to the schedule of a links post a week, and published 3 additional posts: apps for tracking media, making open data a priority (by yours truly), and real life Pomodoros. We also started republishing content on Medium, as a test to see if it gets us more reach.
Speaking of reach, July was the month we had two posts go into Buffer's "Suggested Content" lists. Buffer is a tool for automating posts to social media, which we use ourselves for Exist's account. They also offer a means to "fill up your buffer" with pre-written tweets they've suggested, presumably so folks can keep their social media accounts looking active even if they have nothing of their own to share.
In terms of traffic, this was a boon to us. Our traffic to Exist doubled in July thanks to thousands of tweets of this post, "The difference between productivity and getting things done". It just kept going and going. For two weeks our mentions were full of nothing but this same tweet, over and over. By the end it felt like a sizable percentage of Twitter must have read it. It made me feel good that something Belle had written was so popular.
But what started as something exciting quickly turned sour. The vast majority were tweeting the exact same wording. The number of hits we were getting didn't match the number of tweets we saw. Instead of being read and enjoyed, our content was simply becoming fodder for social media.
Most importantly, I was left feeling dirty, like we had wilfully opted-in to spamming Twitter, with little gain to any party involved. I felt like we were now associated with the sort of "social media strategy" that involves the regular, automated posting of tweets, to links that haven't even been read, like this is a good thing.
It's not a good thing. We won't be involved again if we can help it.
Hustling and housekeeping
This month we met with another local investor who was also uninterested in funding us. After this meeting, Belle and I sat down and had an argument a candid discussion about what we were doing and where we were going. We came to the conclusion that one of our main issues is a lack of focus — we're trying to do everything all at once, and as a result, not doing any of it well. There's not much point trying to find investors if you're unsure of your product strategy. To this end we decided to scrap our search for investors, advisors, or anyone who might want to get us closer to a seed round, and focus on making Exist excellent.
The other main realisation, which is closely related, is that we were kind of just plodding towards our long-term vision, without having yet built something valuable. It's unfair to our paying users to be waiting around for us to get there — we need to build something valuable right now. So that's the plan.
The numbers
Business Stats
- 348 users (299 paid)
- 9% growth from last month with 31 signups
- 9% churn with 33 cancelled accounts
- $891.94 earned after Stripe fees
Ouch :( We need to do something about that churn. Note that our income was slightly higher last month due to a few late backer (yearly account) signups.
Site Stats
- 4,104 uniques to the exist.io landing page (no change)
- 9,326 uniques to the blog (101% increase from June)
- 2,781 uniques to our top blog post, How does your tracker know you're asleep?
- Top traffic sources were Twitter, the Buffer blog, and Entrepreneur.com
Note that despite the huge number of tweets mentioned earlier for the productivity post, it still didn't beat our ever-popular post on sleep tracking.
July: how'd we do?
It felt good to get stuck into new development a bit more. Building the API was fun and remains a source of pride for me, even if nobody is using it (yet). We were one day late on shipping the promised Withings integration, but it did happen, meaning I kept that promise from last month. I didn't deliver on Google Calendar integration, but it's coming.
Churn continues to be a big problem. It hurts a bit less now, like hearing the same insult over and over takes the sting out of it, but it's still something we need to fix. What we hear most is "I like it, but not enough to pay for it."
Next month
This month or next month, to combat churn, we'll most likely roll out a 14 day free trial. We're still working on Google Calendar. We have a few good ideas to bring together all of our data in a more useful way, but this may also take some time to come through. In short, our focus is adding more value — both with new integrations, and making better use of what we have.
Until next month.
Exist helps you make sense of your data and understand your life. You should give it a try.
Author and co-founder @joshsharp is on Twitter and would like to hear nice feedback about this post.