What metric did your last commit improve?
The previous article covered Web Vitals and perceived performance. You can have the fastest app in the world, but if it does not move business metrics, you are only spending time, resources, and money.
A horror story
You receive a ticket, meet every acceptance criterion, write beautiful code, and watch the tests pass. After deploying, you move to the next task. Then leadership raises concerns about acquisition or churn and it feels like another language.
The symptom is clear: you are stuck solving tickets and disconnected from the product. A successful release is not one that breaks nothing; it is one that creates value.
Pirate Metrics (AARRR) for developers
- Acquisition: how users arrive. A 5 MB JavaScript bundle is also an acquisition metric.
- Activation: a long, slow signup makes users leave before seeing value.
- Retention: unstable or confusing software does not earn a return visit. Without retention, there is no business.
- Referral: how easy it is to invite and recommend the product.
- Revenue: checkout problems become real losses.
Three code decisions that affect the business
- Deep linking: Universal Links and App Links turn a discount email into a direct arrival at the right screen.
- JS payload and SEO: every extra KB affects indexing and may force marketing to spend more on Ads.
- Business observability: logging errors is not enough; record intent and context to understand abandonment.
Product Engineer and Growth Engineer
The Product Engineer focuses on Core Value: solving the underlying problem with robust, quality software. Retention is the usual outcome.
The Growth Engineer focuses on Distribution and the funnel: experimentation that reduces friction in acquisition, activation, and conversion.
At a startup you are probably both: one day you build offline support, the next you test a three-step onboarding against a five-step one.
Conclusion
Understanding the business does not take time away from programming. It gives you the judgment to know what is worth programming. The value is understanding how each line of code solves a real problem.