CPI and SPI Explained Without Formula Memorisation
Both indices are simple to calculate and easy to misread, and the misreading costs considerably more than the arithmetic ever saved anybody. The way in is to put the letters aside and ask what question each one answers, because a number whose question you can state is a number you can act on.
What each one is actually asking
The cost index asks how much has been spent on the work actually finished, against what that work was supposed to cost. Above one, the work is costing less than planned. Below one, more. It is a genuinely useful question, and it is honest in one particular way: both halves of it refer to the same completed work, so it compares like with like.
The schedule index asks how much of the work planned to be finished by now has actually been finished. It compares work done against work planned, and the oddity at its centre is that the comparison gets expressed in money, because money is the only common unit available for adding together a survey, a data extract and a fortnight of configuration. Section 2.4 of the PMBOK® Guide Eighth Edition treats both as decision information, which is the right framing. An index is a prompt to look, not a finding.
Three things they do not tell you
The schedule index is not a schedule measure. That is the most important sentence about it and the least often said. It knows nothing about sequence, dependency or the critical path. A project can sit at exactly one, having completed the planned quantity of work, with the two activities everything else waits on both running late, because a large block of non-critical work finished early and the money balances. The index is not lying. It was never measuring what people read it as measuring.
The schedule index always returns to one. When all the work is done, work done equals work planned, whatever happened in between. It grows less informative as a project approaches completion, and across the final fifth of a project it is close to useless, which is exactly the period when people look at it hardest.
Neither index explains anything. A cost index of 0.85 says the work finished so far has cost about eighteen per cent more than planned. It does not say whether that is rework, an estimating error, a rate change, scope that crept in, or one bad month. The number tells you where to spend an hour, and the hour is where the value is.
One point oh one, and three weeks late
A shared services provider was five months into a transition, taking payroll and HR administration from a client across eleven months. The monthly report carried a cost index of 0.92 and a schedule index of 1.01, and the steering group read that the way anybody would: a little over on cost, on track for time.
The schedule index sat at 1.01 because the team had completed, early and efficiently, a large block of process documentation planned for months six and seven. That work was real, it was in the plan, and doing it early moved a substantial amount of planned value into the completed column.
The two things everything else depended on were both late. The client's data extract had been delivered twice and rejected twice for format problems, and the parallel run environment had not been built because it needed the extract. Between them that was three weeks of critical path, and the schedule index could not see any of it, because the money those activities represented was small against the documentation block.
The cost index had noticed, after its fashion. The 0.92 was almost entirely unplanned contractor hours spent on the data extract, which is the same problem appearing honestly in one index and invisibly in the other. Nobody had looked, because 0.92 sat inside the reporting tolerance and 1.01 was reassuring.
What the transition manager changed afterwards was small. The monthly pack kept both indices and gained two lines: the date of the next critical path milestone, and whether it had moved since the last report. Those two lines cost nothing to produce and would have found the problem in month three.
For a PMP® candidate, an index is a question and never an answer. A healthy schedule index sitting beside a description of a late dependency is two pieces of information that mean different things, and the response that fits looks at the critical path rather than at the index. Work out what a figure would look like if the problem described were happening, which is quicker and more durable than memorising what the figure means. A structured PMP exam preparation course rehearses the case where the indices and the narrative disagree.
Keep both indices and add one line beside them: whether the next critical path milestone has moved since the last report. It takes a minute, it is precisely the thing the schedule index cannot tell anybody, and a steering group reading the two together asks better questions than one reading either alone.
Interested in going further?
An index inside tolerance and a critical path three weeks late can sit in the same monthly report without contradicting each other. Omega's PMP® Exam Preparation works through situations where the numbers are correct and the reading is not.
Earned value measures and the rest of the finance work are set out in the PMBOK® Guide Eighth Edition.
Ad · Amazon affiliate link.
References
A123: The PMBOK 8 Finance Performance Domain: What It Really Covers
A128: Do You Still Need to Memorise EVM Formulas for PMP?
A129: Estimate at Completion: What Question Are You Really Answering?
A125: Budget vs Cost Baseline vs Funding
A126: Contingency Reserve vs Management Reserve
PMP and PMBOK are registered marks of the Project Management Institute, Inc.
