View profile

The Hyperoptimisation Paradox 🤖

The Hyperoptimisation Paradox 🤖
By Mustafa Sultan • Issue #45 • View online
Friends,
One of the really cool parts of producing (digital) content is metrics.

On my podcast, not only can I see how many people have listened, but I can go even more granular and see how many people listened to the full episode and where people dropped off.
A well performing episode
A well performing episode
Metrics create a tension however. If I can see which episodes perform well, the next logical step is to study those episodes and try and optimise future episodes to be similar.
But then the tension arrives. Some interviews I really enjoyed — and which I believe are better, perform worse on metrics.
An interview I think is really good, performs worse
An interview I think is really good, performs worse
Every platform gives some similar metric. The Instagram metric is likes. YouTube plays/likes/retention. Twitter favourites/retweets.
Too much focus on metrics, leads to really weird scenarios. For example, A&E 4-hour-wait targets can incentivise hospitals to leave patients waiting in ambulances (as not to trigger the start of the 4 hour wait).
Goodhart’s Law: When a measure becomes a target, it ceases to be a good measure
Through 2020, I put reasonable pressure on myself to 1) release a podcast episode weekly and 2) keep growing my audience.
1) Helped keep me consistent and committed, but also led to a lot of late Sunday nights. It also meant I put out episodes I wasn’t particularly proud of.
2) Meant I spent a lot of time doing things that suck the soul out of me. Like social media marketing.
Now
62 episodes in, I’ve started to question whether these metrics are useful goals — and have changed them slightly.
1) Releasing a weekly episode -> releasing an episode when it suits me, but at least monthly.
2) Growing an audience. Why do I want to grow? I’m consistently getting 300–500 listeners per episode — what will getting bigger achieve? (The quality of guests I can get on is almost completely unrelated to audience size).
I hope that shifting away from vanity metrics will mean that 1) I can carry on podcasting for 10+ years and 2) I can produce a corpus of work I’m really proud of.
Some cool upcoming interviews
Podcast
#062 Sleep and the Evidence Base of Digital Health — Prof Colin Espie (Co Founder Big Health)
All the best,
Musty
Did you enjoy this issue?
Mustafa Sultan

What I'm reading about.

In order to unsubscribe, click here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Powered by Revue