PMs don't "pick stacks" alone—but we do own the product consequences of tech bets. Here's how I'd approach selecting new technology without shiny-object bias.
Questions I'd force first
- What user or operator problem gets cheaper/faster/safer?
- What happens if we wait six months?
- What's the integration and migration cost honestly?
- Who on eng has depth to own failure modes?
- What's the exit plan if the bet sours?
Decision grid I'd use
| Lens | Green flags | Red flags |
|---|---|---|
| Problem fit | Clear constraint removed | Solution looking for pain |
| Ops | Observability + owners | Heroics required |
| Vendor/OSS | Healthy ecosystem | Lock-in with no leverage |
| Timing | Unblocks roadmap now | Resume-driven rewrite |
I'd prototype thin before rewriting the world—and I'd write the decision down.
A Technical PM Certification helps you pressure-test these calls with eng. Keep learning via the newsletter.