SPI (Schedule Performance Index)

How it is computed

SPI = EV ÷ PV. With week-5 planned value at $50k and completed value at $40k, SPI is 0.8 — 80% of the value planned through the status date has been earned. It does not say the work runs at 80% speed, nor by what factor the duration stretches. Being a ratio matters: scale effects drop out, and a trend like "0.9 down to 0.8" reads faster than any list of dates. To see the difference as money instead, use schedule variance: SV = EV − PV, which is −$10k in the example.

Why it matters

It becomes the one-line executive report: the delays and recoveries of hundreds of tasks summarized in a single number. It works as a trend instrument too — an SPI lower than last week's is the signal to descend into the WBS and find what slipped. At the portfolio level — provided baseline quality and progress measurement are comparable — SPI serves as a first-pass screening signal for where to look first. Thresholds make it operational — say, below 0.9 triggers root-cause analysis, below 0.8 demands a recovery plan.

Common misconceptions

Use SPI without knowing its two limits and it will betray you. First, SPI converges to 1.0 near the end: once PV reaches the budget total (BAC) and stops, even a badly late project sees EV catch up and the ratio drift toward 1.0. Late-stage schedule control belongs to remaining-work dates, not SPI. Second, SPI is blind to the critical path: earn value on easy off-path tasks and SPI looks fine while the finish date slips. SPI and the critical path are a pair — read them together.