Business Cases Add Up Gains That Actually Multiply

← Back · ACT II · Chapter 12: Compounded Value

PMI’s Pulse of the Profession 2026, published in May 2026, reports that about one in three complex projects fails, against 13 percent of projects overall, and that complexity “quietly erodes value.” The usual explanation is execution. A more useful one is arithmetic: the value these projects promised was calculated one way and behaves another.

Most business cases are built by addition. A platform is expected to lift three things at once: more qualified leads, a higher win rate, a larger average deal. Each improvement is estimated on its own against the baseline, and the three are summed into a headline number. The sum is tidy and easy to defend. It is also the wrong figure, even when every estimate is fair, because the improvements act on one another.

More leads pass through a higher win rate and close at a larger deal size, so the gains multiply, and the result is larger than the sum. Three 10 percent lifts added together give 30 percent; carried through each other they give about 33 percent. For a moment that reads as good news, since an organization that models it this way sees more value on the table than the additive case showed.

The Same Arithmetic Runs in Reverse

Compounding works in both directions. If leads arrive under plan, the win rate now applies to a smaller base, and the deal size scales a smaller number still. A shortfall in one term pulls down everything downstream of it.

The additive model cannot see this. It recorded three separate lines, so it reports three separate variances and treats the total miss as their sum, when the misses actually work through one another. That is one reason the year-end reconciliation so often surprises people who did nothing careless. Each function reports a believable number: procurement has its saving, sales has its lift, and every line looks survivable on its own. What an additive ledger never shows is that the lines were connected, so the gap between the promised total and the realized one is wider than any single variance explains. World Commerce & Contracting and Ironclad put a figure on a related gap in January 2026, finding that organizations lose an average of 11 percent of contract value after signature. A ledger that treats value as a set of separate lines has a hard time finding where it went.

Time works the same way, which is the subject of A Shortfall in Year One Costs More Than One Year of Value.

A Different Calculation, Run Both Ways

The fix is a calculation that treats the improvements as terms in a product and runs it in both directions. Modeled like that, the number shows its real ceiling and its real floor. Before any governance discussion starts, the organization can see how much more the improvements are worth together, and how much a shortfall in any one of them would take with it.

A vendor whose product does move all three terms gains from a case built this way. The connected model shows a larger prize than the additive one, and when the year-end figure comes in short, it shows which term slipped instead of leaving everyone to argue over the total.

Deloitte’s Finance Trends 2026 survey of 1,326 finance leaders adds a figure that belongs beside this one: 63 percent have fully deployed AI in finance, and only 21 percent report clear, measurable return. Part of that gap is governance. Part of it is a measurement problem that starts earlier, because a number calculated as a sum cannot show how its parts move one another.

References: PMI, Pulse of the Profession 2026: Driving Success in Complex Projects (May 2026); World Commerce & Contracting with Ironclad, “Closing the Procurement Value Gap” (January 2026); Deloitte, Finance Trends 2026 survey release (October 8, 2025).

Keep reading: ACT II · Chapter 12: Compounded Value