What we did wrong in the first 3 months
Three months into creating our coding education app, I want to share with you the main lessons we got out of it, and what we plan to do from now on.
1. Test with focus and test small
The initial mistake was testing out 4 (yes, four) hypotheses at the same time thinking that we had to have everything in place that we thought of in order to get relevant data. An experiment takes time - both to build it and to get the relevant data points. If it flops, the sooner it flops the better for you. If it picks up, you will have time to build up on it.
2. Validate first, perfect next
I am an engineering and product person with a great eye for detail. One of the biggest challenges of having a perfectionist, the detailed oriented background is trying to push everything to be pixel-perfect and work 99% of the time.
It felt uncomfortable to compromise on perfection, but much needed - as it's better to have a product that works on 80% of devices that people want to use, rather than a product that works on 100% of devices that no one wants to use.
3. Be mindful of metrics
If you don't provide daily value to users, don't expect Daily Active Users to be the metric you want to look at. Being a content app very early, it's hard to deliver content daily, so we switched our focus to weekly activity as that is more relevant and feasible for us at this stage.
Weekly activity will give us more insight if the experience is right for the users, and if so, we can optimize the content creation process and deliver value on a daily basis.
Making products that are loved by users it's hard. We are now coming into the 3rd month of full-time work at our education app, and here are the main lessons we have learned.
4. Be patient
Sometimes, things take time. And that's okay. Being patient and observant rather than making rash decisions without enough data on your hands. If you want to move fast, that can be a bit discouraging, but please keep in mind: it's a marathon, not a sprint - in our case, we are building long-term value and both time and effort need to go into it.
5. Don't be afraid to go back to the drawing board
Things, in the beginning very often (at least in our case) don't end up working as planned. Sometimes, the hypothesis is a total miss, so don't be afraid to go back to the drawing board and start from scratch. You are not married to your initial idea - and if it ends up being untrue, please, just let it go.
How we implement these lessons going forward:
-
Our latest experiment took us 5 days to implement and create relevant content - and we started it today
-
We set up the mindful metrics - we are delivering content weekly, and therefore looking at the weekly activity
-
We scraped 3 out of 4 features we initially built, as they fully flopped. Yeah, we spent time on it, and we still like them, but the data told us that the hypotheses behind the features were just wrong.
-
The feature works - it's not perfect, and it's not optimized to the fullest, but it gives users the experience that we want to give them. We will think about scaling and paying off tech debt once we know that, from a product perspective, it works.
And we are patient, with a setup to run the experiment for the next 4 weeks, setting exact metrics that we measure with objectives, and goals. Though, we will be flexible to adapt based on data and user input.
And my final message for everyone still building it: please, think about what you need to do to ship it as soon as today, the most valuable resource that you can get out of it right now is the learning associated with the release.
