Have you got what you paid for?
Comparing budget to spend cannot tell you. Earned value can — and this shows four forecasts with the belief each one rests on, rather than picking one and calling it the answer.
Earned value, with the forecast assumption made visible
Four numbers in. CPI, SPI, variances and four different forecasts out — each with the belief about the future that it rests on.
Where you stand
Four forecasts, and what each one believes
These are not competing answers to one question. They are answers to four different questions about what happens next, and choosing between them is a judgement about your project rather than a calculation.
Every formula used
CV = EV − AC SV = EV − PV
CPI = EV ÷ AC SPI = EV ÷ PV
EAC — four supported methods:
BAC ÷ CPI— the rest costs what the work so far has costAC + (BAC − EV)— the variance was a one-offAC + (BAC − EV) ÷ (CPI × SPI)— being late will cost money to fixAC + (BAC − EV) ÷ (0.8×CPI + 0.2×SPI)— a weighted blend
ETC = EAC − AC VAC = BAC − EAC TCPI = (BAC − EV) ÷ (BAC − AC)
Denominators are guarded. Zero spend or zero planned value is completely normal at the start of a project, and a ratio is reported as unavailable rather than as zero or infinity.
The one idea underneath all of it
Earned value exists to answer a question that comparing budget to spend cannot: have you got what you paid for?
Spending 60% of a budget is meaningless on its own. It is excellent if 80% of the work is done and alarming if 40% is. Earned value is simply the budgeted cost of the work actually completed — what that work was supposed to cost — and comparing it against real spend is what turns a spend figure into information.
Why four forecasts and not one
Every EAC formula encodes a belief about the future, and the differences between them are large. The same project above forecasts anywhere from 120,000 to over 170,000 depending on which you pick.
Dividing the budget by CPI assumes the overrun is systemic — that whatever made the work so far cost more will keep doing so. That is usually the safest default, because cost overruns are more often about estimating and rates than about one bad week.
Adding remaining budget to actual cost assumes the variance is behind you. That is defensible when you can name the cause and it has genuinely finished — a one-time equipment failure, a resolved dispute. It is wishful when you cannot.
The CPI×SPI version assumes lateness costs money to recover, through overtime or extra people. On a deadline-driven project it is often the realistic one, and it is nearly always the most alarming.
A tool that shows one of these and calls it the forecast has made that judgement for you and hidden it.
SPI stops working near the end
Schedule performance index has a flaw worth knowing: it converges to 1.0 as a project completes, no matter how late it is. Once all the work is done, earned value equals planned value by definition, and the index says everything is fine while you are three months over.
SPI is a cost-based proxy for schedule, and it is at its most useful in the middle of a project and its least useful at the end. For schedule, the critical path is the honest instrument.
TCPI is the reality check
To-complete performance index asks what efficiency the remaining work must achieve to land on budget. In the example above it is 1.5: every remaining pound must buy 50% more than the pounds so far have bought.
That number is worth taking seriously. A team running at 0.67 does not suddenly start running at 1.5. When TCPI rises much above about 1.1, the budget is generally gone and the useful conversation is about scope or funding rather than about efficiency.