The quarterly review is nine slides of discipline. Planned value, earned value, actual cost. The schedule performance index reads 0.87, and the PMO reports — correctly — that the program has earned 87 cents of every planned schedule dollar. Then the sponsor asks the only question she came to ask: “So when does Version 4.2 actually ship?”
And the room does the thing rooms do. Someone multiplies remaining duration by 1/SPI on a napkin. Someone else offers, “the baseline still shows June 30.” The honest answer — the one nobody says out loud — is that the metric on the slide doesn’t know.
That isn’t a failure of earned value management. It’s the boundary of it.
What earned value is actually good at
Earned value in agile programs gets a rough reputation it doesn’t fully deserve. EVM earned its place in government and enterprise delivery for a reason: it’s objective, auditable, and brutally consistent. Plan the work, price the plan, measure what you earned against what you planned. Schedule variance and SPI will reliably tell you that you’re behind, and by how much, in units a finance committee can cross-examine. As a governance instrument, that’s exactly the right tool.
But every earned-value number is measured against the baseline drawn on commitment day. And the day your team committed — the day it took its confidence vote — is the best-informed the plan will ever be about itself, and the least-informed it will ever be about what happens next.
The baseline doesn’t move. The world does.
After the vote, post-commitment erosion goes to work. This is why release dates slip after PI planning: scope creep, dependencies surfaced too late, and capacity changes. EVM feels all three. It just feels them late, and without attribution.
- Scope creep changes the denominator. Approved change requests re-price planned value; unapproved ones quietly make “100% complete” a moving target. SPI dips — but it can’t tell you which change did it.
- A dependency surfaced too late shows up as work that can’t be earned — tasks parked at 0% while the calendar burns. The variance appears weeks after the blocking decision actually happened.
- Capacity changes bend the earn rate. A departure, a loan-out to another program, a hard flu season — the index degrades gradually, and the explanation lives in someone’s head, not in the metric.
By the time schedule variance crosses a governance threshold, the erosion is old. You’re reading last month’s weather.
From “how far behind” to “when, and how sure”
Here’s the pivot. Your team’s commitment was never a napkin number. A PI confidence vote is expert engineering judgment — the people closest to the work weighing scope, dependencies, and capacity, and promising a date on evidence they validated. The judgment was never the problem. The problem is that the vote is a snapshot, and the assumptions underneath it start aging the moment the room empties.
Release Management, Roadmaps & Product Portfolio for Jira keeps that judgment live. The team still owns the assumptions — which throughput window to trust, whose capacity counts, what’s in scope. The app does the arithmetic at Monte Carlo scale over the team’s real Jira throughput, continuously, and carries the committed date with its confidence band: the 50%, 85%, and 95% dates, refreshed as reality arrives.
The earned-value questions translate cleanly:
- Schedule variance becomes drift in days — the live gap between the committed date and the date at the confidence level your team chose to promise against.
- SPI 0.87 becomes a sentence you can act on: “the committed date now sits below your 85% line — here are the days, and here’s whether scope, dependencies, or capacity moved them.”
- The estimate-at-complete becomes a forecast you can defend, drawn from throughput the team actually demonstrated rather than a ratio projected forward.
Variance you can renegotiate, not just report
The point of seeing drift in days, attributed to a cause, is that you can act while options are still open. If scope moved the date, it’s a cut-line conversation. If capacity moved it, it’s a staffing or calendar conversation. Teams renegotiate before the slip — not after the variance report makes it official. Because re-baselining makes a slip official; it doesn’t make it small. And unlike a say-do ratio, which scores the commitment after the PI ends, a live band warns you while the commitment can still be kept.
So keep EVM. It’s the governance layer your program owes its sponsors. Give it the companion it’s missing: a forward-looking committed date that answers the sponsor’s actual question — when, and how sure. You can see how variance becomes a live reading in the docs.
The sponsor’s question, answered on the spot
Next quarterly review, the SPI slide keeps its place. But when the “when does it land” question comes, put the band next to it: “Committed: June 30. Currently 82% confidence — four days of drift, three of them from the scope change we approved on the 12th. Here’s the renegotiation we propose.” That’s earned value doing what it does best — and your team’s judgment, kept current, doing the rest.
Ready to answer “when” with confidence? Put your committed date on a live confidence band with Release Management, Roadmaps & Product Portfolio for Jira.
Leave a Reply
Your email is safe with us.