We have a legacy auth service inside our payments flow. Incident hours every month, slower releases on anything touching it.
I ran payback math, took it to planning, still got zero headcount for non-roadmap work. CISQ puts poor quality at $2.41 trillion a year in the US. Helped my case exactly none.
Has anyone actually gotten remediation funded on ROI math? What made finance believe it?
zero headcount for non-roadmap work
Finance funds bleeding deals. Not math. Bring one.
Not math is wrong. I watched math win once.
We had a session service paging us twice a week. Every page woke two people. On-call was ready to quit, so I wrote one page and took it to finance. Six months of page counts, hours burned, plus three pen-test findings priced as a failed audit. One number at the bottom: $61k a year to do nothing.
Finance funded two sprints. The thing that sold it was the page log. Real timestamps. They could see the bleed.
Tried the same one-pager on billing last year. Same format, worse numbers. Got nothing. Still don’t know why that one died, and nobody on that team can explain it either.