Do we need Backstage to improve DevEx?

Not to start. Backstage can be useful as an index over your paved road, but it won’t cut wait times by itself. Ship fast CI, preview envs, and policy-as-code first; add a portal after the metrics move.

Can we do this in a regulated (PCI/SOC2/HIPAA) environment?

Yes. Put policy in code (OPA), require reviews where risk is real (infra, secrets, PII), and make every change auditable via GitOps. Automated checks usually strengthen compliance and shorten audits.

Do we need a monorepo to get these gains?

No. The biggest gains come from CI caching, preview envs, and clear ownership. Monorepos can help with dependency control, but they are neither necessary nor sufficient for speed.

What if our tests are the bottleneck?

Split unit vs integration, parallelize, and cache. Quarantine flakes and fail fast. Tools like `pytest-xdist`, `jest --runInBand` vs `--maxWorkers`, and test sharding in CI cut time without rewriting tests.

How do we keep teams from ignoring the paved road?

Make it the fastest option. Measure adoption, deprecate snowflakes, and require a written exception with SLO impact. Celebrate teams that hit the SLOs on the paved road.

Platform-productivity · Oct 16, 2025 · 10 minute read

Stop Timing Standups. Start Timing Waits: Measuring Friction and Killing Hand‑Offs with a Paved Road

If your PRs age like cheese, you don’t need another platform — you need a stopwatch, a paved road, and the courage to kill bespoke snowflakes.

Alex Kim

Principal, Platform Engineering, GitPlumbers

20 years building and rescuing platforms at scale (AWS, Atlassian, two unicorns you bank with). I’ve led teams through microservices rewrites, SOX/PCI audits, and three CI/CD overhauls that actually moved DORA metrics.

You don’t need a developer portal. You need a stopwatch.

Back to all posts

Stop Timing Standups. Start Timing Waits: Measuring Friction and Killing Hand‑Offs with a Paved Road

Key takeaways

Implementation checklist