B2B logistics · Growth Engineer Lead · 2025
Shipping product features from the growth team
Context
Most of the value sat in small things that never made the roadmap. Growth took those on directly.
Trigger
Trigger
Experiments kept dying in a queue longer than the experiments themselves.
The build
- 01
Small in-product surfaces
Marketing banners and prompts inside the product, scoped narrowly enough to ship and switch off safely.
- 02
Marketing site
A Next.js marketing site owned by growth, so pages ship the same week they are written.
- 03
Testing
A/B tests through VWO or direct script injection, depending on what the surface allowed.
- 04
Dashboards
Internal dashboards so the teams asking for the features could see whether they worked.
Stack
- Next.js
- React
- VWO
- SQL
- Metabase
Result
6 WKS
typical lead time removed from growth experiments
Honestly
What I'd do differently
We never wrote down what growth was allowed to ship versus what belonged to product engineering. That was fine until two teams touched the same component in the same week. A one-page rule — growth owns marketing surfaces, banners and the marketing site; engineering owns core product code — would have prevented the collisions and the awkward conversations after them.