To-Complete Performance Index

Cost technique · See it on the map

Runnable here

The cost efficiency remaining work must hit to land on a realistic target cost.

When to use it

Any time CPI is off 1.0 and someone's asking whether the project can still hit its original budget. Use TCPI against BAC — (BAC − EV) / (BAC − AC) — to answer 'can we still make the original number', and TCPI against a revised EAC — (BAC − EV) / (EAC − AC) — to check whether even that revised, more forgiving forecast is achievable at the efficiency the team has actually been running at.

When to avoid it

Don't present a TCPI-against-BAC figure as a plan without comparing it against the CPI the project has actually been achieving — a TCPI far above current CPI isn't a target, it's a number nobody has a believable way to hit, and presenting it as achievable is arithmetic fiction dressed up as a forecast.

Steps

1. Get current EV, AC, and BAC (or a revised EAC, if one exists and is trusted).
2. Decide which denominator answers the question being asked: BAC if the question is 'can the original budget still be hit', EAC if the question is 'is even the revised forecast realistic'.
3. Compute TCPI = (BAC − EV) / (BAC − AC), or (BAC − EV) / (EAC − AC) for the EAC form.
4. Compare the result against the CPI the project has actually achieved to date.
5. If TCPI sits well above achieved CPI, treat that gap as the finding — it means the remaining work needs to run meaningfully more efficiently than the work already done, and say so explicitly rather than presenting the target alone.

What it produces

• A required-efficiency figure (TCPI) for the remaining work, against either BAC or a revised EAC.
• An explicit comparison of that required efficiency against what the project has actually been achieving.

Common pitfalls

• Reporting TCPI against BAC without ever computing or checking it against achieved CPI — the gap between the two is the entire point.
• Continuing to use the BAC form after everyone has effectively accepted a revised EAC, so the TCPI reported no longer answers a question anyone is actually asking.
• Treating a TCPI just modestly above current CPI as fine, without asking what specifically will change about how the remaining work is run to close even a small gap.

Worked example

On the same community-center renovation (BAC $850,000, EV $348,500, AC $410,000, CPI 0.85), TCPI against BAC is (850,000 − 348,500) / (850,000 − 410,000) = 501,500 / 440,000 ≈ 1.14 — the remaining work would need to run at 1.14 efficiency, well above the 0.85 the project has actually delivered so far, to still hit the original $850,000. Against the earlier EAC of $1,000,000 instead, TCPI = (850,000 − 348,500) / (1,000,000 − 410,000) = 501,500 / 590,000 ≈ 0.85 — exactly the efficiency already being achieved, meaning the revised forecast is the believable one and the original budget, at this point, is not.

Source

• PMBOK-6 §7.4.2.2

Where it comes from: this technique is named by the PMBOK Guide, 6th edition.