Situation
Early in 2024, at the CTO’s request, my manager developed a theory: engineering time and effort were being spent in ways inconsistent with the strategic product roadmap, the prioritized feature-delivery list, and investment in development to drive revenue.
I had seen the same pattern at Nike and again running Cloud Ops at SentinelOne, where allocation decisions sat between Engineering and Product Management and neither had real visibility into how engineering effort was actually being allocated.
Two separate Jira instances, one on-premise and one in the cloud, with highly divergent structures for artifacts, issues and reporting, made unified reporting hard.
Objective
Give leadership real visibility into where engineering effort went, align it to roadmap priorities, and sponsor the initiatives that followed.
Approach
- Buy, then build. I first recommended buying a DevOps / engineering-delivery solution and assessed several. There was no budget, so I built my own in Google Sheets: an individual-contributor report and a management report, fed from Jira via API, mapping work from the bottom up to fiscal-year priorities.
- Past the Sheet limits. Jira held millions of artifacts, far past what a Sheet can hold, so I built a custom CSV and Google Drive persistence layer with scheduled jobs to extract, persist and process the data.
- One platform. In parallel I pushed the unification of process, artifacts and platform into Jira Cloud, and then integration with BigQuery. By late summer and fall 2024 everyone was on a common set, and we routinely showed work attribution up the full hierarchy of initiatives.
- Standards. We standardized story-point assignment and made unplanned work such as remediation visible. That pushed unsponsored work toward a low target and made the case for SRE and QA functions to absorb interrupting unplanned work.
- Bring the dashboard to them. By late 2024 leaders could see delivery against initiatives but rarely logged in. In Apps Script I built automated scorecard delivery drawing on Jira, BigQuery, GitHub Enterprise and Workday. Within 12 hours of the end of each sprint and each 13-week release, six 2-week sprints plus a 1-week interim, every engineering leader received a personalized Google Sheet by email: a summary of delivery against objectives plus a dashboard for themselves, each leader beneath them and any direct-report ICs.
- Capitalization and depreciation. The same platform revived a 2021 initiative of mine: better reporting for capitalizing and depreciating engineering effort. I attributed work hours down to the individual IC using a clean unit of 1 story point = 1 average engineer-day, with a daily-updated current-quarter forward projection and a basis for longer-range projections.
Net results
- Engineering behavior changed at scale, because the scorecards fed semi-annual performance reviews and staffing, reassignment and hiring decisions. Unsponsored work dropped toward its target and effort re-aligned to roadmap priorities.
- A meaningful impact on time-to-market and on pointing engineering resources at revenue generation.
- The prototype was promoted to production as the org’s primary source of engineering-delivery visibility.
- Better attribution of real engineering effort produced a meaningful increase in capitalizable spend and strengthened audit-ready financial reporting.
- I operated the delivery-reporting platform and the capitalization and depreciation reporting from 2024 through 2026.
Takeaway: deliver the scorecard to the inbox and tie it to reviews. Visibility that is pushed and consequential changes behavior; visibility that is merely available does not.
Scope and technologies
Jira on-prem and cloud via API; Google Sheets and Apps Script; a custom CSV and Google Drive persistence layer with scheduled jobs for millions of Jira artifacts; Jira Cloud consolidation; BigQuery, GitHub Enterprise and Workday integration; automated per-leader scorecard email delivery; story-point standardization; labor-cost-to-revenue correlation.