NPV, IRR and payback, explained for operators

You run operations, or customer experience, or a transformation programme. You are not a finance person. But the moment your business case reaches a CFO, it will be judged in finance's language: payback, NPV, IRR. Here is what each one actually measures, where each one lies to you, and why looking at more than one year changes decisions.

Payback: how fast do I get my money back?

Payback period is the simplest of the three: how long until cumulative benefits equal the upfront cost. Spend £100k, save £50k a year, payback is two years.

Its virtue is that everyone understands it. Its vice is everything else. Payback ignores whatever happens after the break-even point — a project that pays back in 18 months and then dies scores better than one that pays back in 30 months and then compounds for a decade. It also ignores the time value of money: £50k arriving in year three is treated as if it were worth £50k today.

Payback estimates are also routinely optimistic. In one of the few longitudinal data points available, executives implementing intelligent automation reported average pilot payback lengthening from 16 months to 22 months between surveys (Deloitte, 2022) — real-world payback drifted a full six months beyond what the same population had reported earlier. If your case hinges on a precise payback month, it is fragile.

Use payback as a communication device and a risk screen — shorter payback means less time exposed to things changing — never as the deciding number.

NPV: what is this worth in today's money?

Net present value fixes payback's blind spots. It adds up every year's cash flows over the project's life, but discounts each future year, because money later is worth less than money now — it could have been earning a return elsewhere, and it might not arrive at all.

The discount rate is the honesty dial. A typical corporate rate of 8–12% says: a pound in year three is worth roughly 75–80p today. Sum the discounted benefits, subtract the discounted costs, and you get one number in today's money. Positive NPV means the project creates value at that discount rate; negative means it destroys it.

NPV is the measure finance trusts most, and it is worth knowing that this is also how academics define customer value itself — the peer-reviewed customer lifetime value literature describes a customer's worth as "the expected sum of discounted future earnings" (Gupta, Lehmann & Stuart, 2004). Discounting is not a finance affectation. It is how value is defined.

IRR: what rate of return does this project earn?

Internal rate of return answers a different question: if this project were an investment product, what annual return would it be paying? Technically, it is the discount rate at which the NPV falls to exactly zero.

IRR's appeal is comparability. "This returns 34% a year" can be set against other projects, or against the company's cost of capital. If IRR is comfortably above your discount rate, the project clears the bar with room to spare.

Its trap is that a percentage hides scale. A tiny project can post a spectacular IRR while adding trivial value; a large project with a modest IRR may be worth far more in absolute terms. Read IRR and NPV together: IRR for "is the return rate attractive?", NPV for "how much value, in money?".

An illustrative example

The following is an illustrative example — invented numbers, chosen to show the mechanics.

A self-service project costs £120k upfront and delivers net benefits of £60k in year one, £75k in year two, £75k in year three as adoption builds.

• Payback: cumulative benefits pass £120k during year two — payback is exactly two years.
• NPV at a 10% discount rate: year one's £60k is worth £54.5k today; year two's £75k is worth £62.0k; year three's £75k is worth £56.3k. Total discounted benefits: £172.8k. NPV = £172.8k − £120k = about £52.8k.
• IRR: the discount rate that zeroes the NPV works out to roughly 32% — well above the 10% hurdle.

Now the payoff of the multi-year view. Judged on payback alone, this project ties with any rival that also breaks even at month 24 — including one whose benefits stop dead at that point. Judged on NPV, they are nothing alike: the rival's year-three £75k is the difference between roughly £52.8k of value and near zero. Same payback, completely different decisions. That is the case for never presenting a one-year number when the benefits genuinely run for several years — and for being explicit about how many years you are claiming, since benefit durability is an assumption a reviewer is entitled to challenge.

What to actually put in your business case

• Lead with NPV over an explicit horizon (three or five years, stated), at a discount rate agreed with finance — asking them for the rate is itself a credibility move.
• Show IRR alongside, so the return clears or fails the hurdle visibly.
• Include payback for the audience that thinks in it, but frame it as a risk measure.
• Run all three on your conservative scenario, not just the moderate one. A project with positive NPV in the conservative case is close to unarguable.

---

LeadersToolset builds this in rather than leaving it to your spreadsheet: every calculator projects conservative, moderate and optimistic scenarios, and the multi-year financial panel computes NPV and IRR over your chosen horizon with every assumption listed in an audit trail. See how it hangs together in the sample report or the methodology, then choose a calculator and put your own numbers through it — the first calculator is free, with unlimited re-runs.