What was happening
The Predictive Index was losing time to outages. Availability sat around 92%, development cycles were slow, and there was no technical strategy. The codebase was a flaky monolith, largely built by an offshore agency years earlier, and the engineering team maintaining it was undersized.
At the same time, the company was moving from a long history of sales-led growth to product-led growth. That depends on a product that stays up.
The call
The fix was not one thing. I led a stability initiative across CI/CD, infrastructure, incident management, test process and tooling, and the engineering culture around all of it. We went from monthly deploys to continuous delivery, and started running 5-why post-mortems so the same failure did not happen twice.
I overhauled planning with OKRs and strategy briefs, worked with finance on the org strategy, and set the hiring plan.
What happened
Availability went from about 92% to 99.9%. Engineering more than doubled in under a year.
We launched the company’s first machine learning product, which recommends actions to managers and opened a valuable new sales conversation. We built the platform for monetization, value metrics, sharing and a free-tier workflow: the pieces the product-led strategy needed.
The multi-year platform strategy I set kept running after I left.
What I'd tell a founder
Reliability is a growth problem. A product-led company cannot grow on a product that goes down.