User engagement is just another word for user complaints
Description
A four-panel comic from 'WHYARENTYOUCODING.COM' featuring two developers sitting at their laptops. In the first panel, one developer expresses frustration: 'WHAT'S THE POINT? EVERY TIME WE MAKE CHANGES TO OUR WEBSITE OUR USERS JUST COMPLAIN.' In the second panel, his colleague offers a motivational spin: 'USERS COMPLAIN BECAUSE THEY CARE, AND THEY CARE BECAUSE THEY GENUINELY LOVE THE APP. IF ANYTHING, YOU SHOULD BE PROUD OF THE NUMBER OF COMPLAINTS.' After a moment of thought in the third panel, the first developer, having completely missed the point, eagerly asks in the fourth panel: 'HEY, CAN I WORK ON THE NEW POPUP ADS? I BET THOSE WILL GENERATE LOADS OF COMPLAINTS!' His colleague dryly responds, 'THAT'S THE SPIRIT.' The comic satirizes corporate attempts to reframe negative feedback as a positive metric. It highlights the dangerous, yet humorous, misunderstandings that can arise when developers are encouraged to view user complaints as a sign of success, leading to a logical conclusion of intentionally creating user-hostile features
Comments
12Comment deleted
The PM told me to optimize for user engagement. I'm just following the data that says nothing gets a user to click more than trying to find the 'X' on a popup ad
We set an OKR to “increase user feedback per session,” and now the team’s A/B-testing which popup gets cursed at faster - Goodhart’s Law has a pull request
After 15 years in tech, I've learned that the best way to validate your monitoring dashboard is working correctly is to deploy popup ads - if the complaint metrics don't spike within 30 seconds, you know your observability stack is broken
This perfectly captures Goodhart's Law in action: 'When a measure becomes a target, it ceases to be a good measure.' The moment you optimize for complaint volume as a proxy for user engagement, you've created a perverse incentive structure that would make any growth hacker proud - and any UX designer weep. It's the enterprise equivalent of training your ML model on the test set: technically you're hitting your metrics, but you've fundamentally misunderstood the assignment. Next sprint: implement dark patterns, break the back button, and auto-play videos with sound. KPIs through the roof, NPS through the floor - perfectly balanced, as all things shouldn't be
Treat complaints as engagement and watch Growth A/B test popups for max CPM - complaints per minute - while SRE labels it “synthetic load” on the support queue
User complaints are the ultimate PMF signal: silence means forgotten legacy, rage means you're shipping fast enough to annoy power users
Goodhart's Law in action: make 'complaints' the OKR and someone will A/B test a full-screen modal where the close button is 1px, hitting the target at scale
oh god😂😂😂😂😂 Comment deleted
fuck popup ads Comment deleted
niggers Comment deleted
Ow. Comment deleted
* pop up blocker entered the chat * Comment deleted