By Max Sherman ยท Oct 8, 2024
We hired a new engineer, growing the team by 50%. In their first week they were able to propose and implement a better solution to a problem that we had around proxying requests to our server to avoid corporate firewalls from breaking our app. I'm super excited about it.
The bottleneck of the business is speed of engineering. We need to iterate faster to make the product something people want, faster, more reliable, and useful. Putting more energy into engineering is going to accelerate our growth.
As we grow the team, I want to make sure we are executing on only high priority tasks. What does it mean to be high priority? The highest priority tasks are the ones which, when done, yield the largest increase in conversion rate, and also require the minimum amount of time to complete.
That is a working definition but I think it's ok. It optimizes for quickly increasing the conversion rate.
One issue is that it's not always obvious which tasks will increase conversion rates, or how long those tasks will take. In practice, you have to make educated guesses. Part of me feels like prioritization is a kind of boogey man, when the real answer is to just do work as fast as possible and learn along the way.
To try and uncover more information around what should be prioritized, I would like to build out an automated way of collecting feedback from users. I think eventually we can convert this into a system to automatically get testimonials.
Some known areas of improvement are still reliablity and performance. Improvements there will always translate to a better user experience, and I believe a higher conversion rate.
We ran a couple small A/B tests with the paywall, though nothing has had a major effect. Low samples makes it hard to get reliable signal, you have to wait longer to collect more data. I think it would be worthwhile to A/B the headline and subheadline copy on top of the paywall.
Some iteration is needed on all of these, but super happy to be shipping these.