EDLP vs. High-Low: A Tie on Paper, $82K Apart in Practice

Run Cinderhaven Provisions' hot sauce line through a full year of high-low pricing and it nets $3,974,400. Run the same line at a flat everyday-low dead-net and it nets $3,974,400. The two scenarios tie to the dollar, because that is how a buyer frames the choice: same money, different shape. Cinderhaven is a fictional company and every figure here is a synthetic dataset; the two price structures are the ones every specialty brand is asked to choose between. The tie is real. So is the $82,067 a year it costs to run one side of it.

That $82K never appears in the pricing conversation, because none of it is price. It is reconciliation labor, double-billed allowances, forward-bought inventory, and fill-rate misses on demand spikes the promotion itself created. A brand that walks into the review with only the blended-price math has built half a model. The buyer has built the other half, and it is the half that decides.

Two price models, one net number

The line: 45,000 cases a year, twelve jars to a case, at a $96 everyday list, which is $8 a jar wholesale. High-low runs four promo windows of three weeks each at $19.20 off per case, a 20% deal funded off-invoice and through scan allowances. Promo weeks move 1,500 cases against a 675-case baseline, so twelve deal weeks carry 18,000 cases, 40% of the year's volume. The funding bill is 18,000 × $19.20 = $345,600, and net revenue lands at $4,320,000 − $345,600 = $3,974,400. Blended net per case: $88.32.

The everyday-low scenario is the dead-net ask that comes from Walmart or Costco, which do not charge slotting and instead want the trade cost baked into the price: price at the blended figure from the start, no deal calendar. $88.32 × 45,000 cases = $3,974,400. Identical.

Holding annual volume equal across the two scenarios is the conservative assumption, and it has data behind it. In the classic field experiments at an 86-store grocery chain, Hoch, Drèze, and Purk found that a 10% cut in everyday price bought only a 3% gain in volume. Everyday price is a blunt demand lever. Promotions spike volume during the window, but much of the spike is the same shoppers buying early. The price sheet says the choice is free. The ledger disagrees.

High-low pays a second time, in paper

Under EDLP, the discount leaves once, as a lower invoice. Under high-low, $345,600 of funding leaves as paper: off-invoice lines to reconcile against POs, manufacturer chargebacks billed after the window closes, scan allowances deducted from remittances at each retailer's own pace. Trade dollars are a reduction of the transaction price, not a marketing expense, and price reductions that arrive as deductions have to be matched to the events that authorized them.

For Cinderhaven's four windows, that is roughly 640 promo-linked rows in the deduction ledger for this one line: four events across six EDI partners and the regionals, each throwing off-invoice, scan, and chargeback lines for a dozen weeks. At fifteen minutes a row to match, validate, and clear, the year costs 160 hours of staff time, about $8,800 at a loaded rate of $55 an hour. Matching is also where the double-dips surface: a scan allowance billed against units an off-invoice deal already covered. At a 2% double-billing rate, $6,912 of the $345,600 goes out twice, and it stays gone unless someone catches it. This is the stretch of the trade promotion process where execution breaks: the deal was approved in a sales meeting and settled in an accounts-receivable queue.

EDLP writes the discount into the price once. High-low writes it 640 times and asks the brand to check the math.

Part of the deal buys cases that were already sold

A distributor facing a three-week window at $19.20 off does what any rational buyer does: loads up. If 10% of promo volume is forward-buy, inventory bought at deal price to sell after the window at full margin, then 1,800 cases × $19.20 = $34,560 of funding produced no incremental retail movement at all. The discount subsidized the distributor's margin on ordinary sales.

The spike costs a second way. Promo weeks run at more than twice baseline, and a production schedule tuned to 675 cases a week strains at 1,500. If deal weeks ship at 97% fill instead of 99%, the line misses 360 cases at $88.32 net, $31,795 in revenue that walks, before any fill-rate fine lands on top. The promotion creates the demand spike, then bills the brand for missing it.

High-low needs 930 incremental cases to earn its keep

Sum the carry costs: $8,800 in reconciliation labor + $6,912 in double-dip exposure + $34,560 in forward-buy leakage + $31,795 in spike-week short-ships = $82,067 a year on a $4.3M line. To cover that, the four windows must generate 930 cases at $88.32, about 11,000 jars, that would not have sold otherwise: 2.1% of annual volume in incremental demand, net of pull-forward and pantry-loading.

That bar is set generously, because it values an incremental case at its full net price. Price it at a 35% contribution margin instead and the windows need about 2,650 incremental cases, 5.9% of the year. The model has to say which it is using; the price sheet never does.

Some promotions clear either bar. Most do not. In a 2014 Nielsen review of 39 million promotional events totaling $555 billion in U.S. retail sales, almost three-quarters of promotions did not break even. The burden of proof sits on the deal calendar, not the dead-net.

The buyer's enthusiasm is not evidence either way. The same Hoch experiments found EDLP cut the retailer's profit by 18% while high-low raised it by 15%: the retailer has arithmetic reasons to prefer the promotion cycle that have nothing to do with whether it works for the brand. When a buyer offers the choice, both scenarios are already modeled on their side of the table. The only question is whether the brand shows up with its own.

Bring the promo calendar

Send me last year's promo calendar for one line and the deduction ledger that goes with it. I will run both scenarios on your numbers and write back with which windows earned their funding and which ones paid the distributor to buy what it was already going to buy. Start with one line's calendar. The tie on paper usually survives the exercise; the $82K usually does not.