The number is hiding in the shadow

An engineer working on a major automotive app asked me this last week: my best work is qualitative, I fixed a broken auth flow, I improved the architecture, so how do I put numbers on it? Everything I read says quantify your impact, and I have nothing to count.

Almost every engineer has this problem, and the standard advice, add metrics, is useless without the method. So here is the method.

When the change is qualitative, you do not quantify the change. You quantify its shadow.

Every real improvement casts a measurable shadow somewhere. The auth flow you fixed: its shadow is support tickets about login, drop-off between the login steps, session re-authentication frequency. The refactor you shipped: its shadow is how long features in that module took to build before and after, or how many bugs that area produced per quarter. The review culture you improved: its shadow is defects caught before production versus after.

And here is the part nobody tells you: you can often pull the before from history after the fact. Ticket counts, crash dashboards, analytics funnels, sprint records, they were all recording while you worked. You did not need to plan the measurement to harvest it.

When even the shadow has no number, quantify the scope instead: every user who logs in, both platforms, every session, forty engineers who touch that module. Fixed the auth flow becomes removed a failure path affecting every login across both platforms, support tickets on sign-in dropped the following month. Same work. Committee-readable.

This week's fix, 20 minutes. Take the one piece of work you are proudest of from the last year. Write down what it touched, then hunt its shadow: one metric that existed before and after, from any dashboard, ticket system, or sprint history you can still access. If no metric survives, write the scope sentence instead. That one line goes on your CV, in your next review, and in your next interview answer, and it is the difference between I improved the architecture and evidence.

Your work already cast the shadow. Most engineers just never go back to measure it.

See you next Tuesday.