One song at a time

I’ve recently shut down a couple of projects:

  • Listkeeper - After telling the whole story of the app in my last post, I made short work of it: I deleted all CloudFormation stacks, S3 buckets, DynamoDB tables, and CloudWatch logs. As a final act, I closed the AWS account dedicated to the project and archived the GitHub repo.

  • dilbert-feed - The Dilbert comic strip was discontinued in March 2023. I considered bringing back a version of the feed that acts as a time machine and delivers comics from the past but ultimately decided against it. (Dilbert’s creator being racist made it easy for me.)

  • whatcdkisit.com - I disabled auto-renewal for ‌the domain in DNSimple. I may still host the website elsewhere – it’s a neat little playground after all – but I currently lean towards closing it down as well.

All three are highly automated serverless projects powered by managed services such as AWS Lambda and AppSync. Thanks to the serverless model, the total cost of running these applications – including infrastructure, development, and maintenance – is comparably low.1

So why would I shut them down?

For a start, the sheer lack of user traction has undoubtedly influenced my decision. Maintaining motivation is hard, doing so without users is harder still.

More essentially, though, I need to focus. Focus on what is truly important to me and eliminate the rest.

Although none of the projects cost me money, they continue to command attention: the time I spend thinking about them, stressing about not having done X yet (deploying security patches, etc.), or noodling around with inconsequential implementation details despite knowing better.

Those little things add up, serverless or not. I can only muster a limited amount of energy before running out of steam.

The antidote is doing less but better, so I can make the highest possible contribution toward the goals that matter to me.2 My ultimate goal is to build a successful product business.

I want to end this short post with an apt analogy. In one of the first episodes of Hackers Incorporated, Adam and Ben talk about the idea of running your business like a band by doing one thing – working on one song – at a time. I believe that’s a wonderful idea for small teams and solo creators alike. :metal:

  1. Yep, I’m linking to a Deloitte whitepaper. It’s quite good. ↩

  2. I highly recommend Greg McKeown’s book Essentialism. ↩