59% of trade promotions lose money. Can you tell?

Roughly 59% of trade promotions lose money worldwide, 72% in the US. The uncomfortable part isn't the hit rate — it's that most brewers and CPGs can't say which of their own promotions are in that 59% until the quarter closes, if then. A live promo P&L needs order, redemption and cost data joined within days, not months; in most estates those three signals live in three systems that never meet. It's an architecture problem in a finance costume.

What does a money-losing promotion actually mean?

A promotion loses money when the incremental margin it generates is smaller than the discount, funding and execution cost poured into it. That sounds trivial to check, and it isn't. The word doing the work is incremental: volume you would not have sold anyway, measured against an honest baseline. Forward-buying, pantry-loading and cannibalisation of the full-price line all inflate the sell-in number that makes a deal look like a win. A ten-percent volume bump on a mechanic that simply pulled next month's orders into this month is not growth; it's a loan you repay with interest, and it books as a success.

To grade a promotion you therefore need two things at the same time: incremental volume attributed to a specific mechanic, and the actual money spent against that mechanic — list discount, off-invoice allowance, scan-back, listing fee, the lot. Most organisations can produce the first as a rough sell-in lift and the second as a month-end accrual, but almost none can put the two next to each other, matched to the same promotion, while it is running. So the profitability question goes quietly unanswered, and the answer defaults to optimism.

Why is the number big enough to lose sleep over?

Trade spend runs at 20 to 27 percent of revenue — the second-largest line on the P&L after cost of goods, around $500 billion a year globally and over $200 billion in the US, as of October 2025. That is not a marketing sundry; it is the largest discretionary lever most beverage businesses own. When a coin-flip share of that spend destroys value, the leakage isn't a rounding error on a side budget — it is happening on the second-biggest number in the accounts. For a top-3 global brewer, a single point of trade-spend efficiency is a serious sum, and right now it is being decided by promotions nobody can grade until they are already over.

Why can't you tell which promotions lose money?

Because the measurement depends on data that has never been assembled in one place. The Promotion Optimization Institute's State of the Industry work reports that more than 70 percent of small and mid-sized CPGs have limited visibility into promotion performance, and over half struggle to harmonise data across their trade partners. Large brewers with mature planning tools are not exempt; they hit the same wall from the other side, because the harmonisation problem is structural, not a missing feature you can license your way around.

Every distributor, wholesaler and key account reports redemptions and claims in its own format, on its own clock. Every operating company then solves the joining problem locally, which is exactly why every OpCo rollout restarts from zero and why the resulting numbers refuse to line up across a region. You end up with dozens of slightly different definitions of incremental and spend, none of which reconcile, and a head-office deck that averages them into a figure no one fully trusts. A trade-promotion management tool doesn't rescue this on its own: it plans against the same disconnected sources, so even a mature planning layer still produces a post-mortem rather than a live view.

Where does the money actually hide?

It hides in the gaps between three signals that each live in a different system and move at a different speed. Promo profitability is the join of all three, and a join is only ever as fresh as its slowest input.

Read the table as a timing problem, not a data-quality one. Order data is quick; redemption claims trickle in over weeks as partners submit, dispute and resubmit; funding sits as an accrual in the SAP financial core until month-end, booked by ledger account rather than by mechanic. By the time all three reconcile, the promotion has finished and the next one is already live. It is the same reconciliation discipline that a returnables ledger that actually reconciles demands, and the same accrual-versus-redemption timing trap that makes loyalty points a liability your CFO sees first — three faces of one problem: money committed now, settled later, matched never.

So what would actually change the picture?

Not another dashboard stacked on the same disconnected sources; that only renders the lag more attractively. The change is putting order, redemption and cost events on one event stream so a promo P&L is computed while the promotion runs, not reconstructed after it — an architecture decision far more than an analytics one, and the direction the rest of this diagnosis points toward.

The brewers who pull ahead won't be the ones with the cleverest promo-optimisation model; they'll be the ones who can connect an order captured this morning to the funding it drew down before a rep plans next month's deal. Until order, redemption and cost share a single clock, promotion analytics stays a forensic exercise — an autopsy rather than a vital sign. The work worth doing over the next few years is moving that measurement from after to during, so the 59% stops being a statistic you read and starts being a list you can act on.