Full-price sales are one of the metrics a brand is actively trying to win on, precisely because relying on markdown to close the gap has become the default rather than the exception. Whether brand awareness is strong enough on its own to carry sales outside a handful of major sale moments, or whether the business has quietly become dependent on discounting to get there, often stays unknown. If markdown has been doing more of the work than awareness for a while, that is a finding the reporting function has never surfaced, which makes the eventual discovery worse than the fact itself.
A read on how much of outside-sale-period revenue is genuinely full price versus propped up by smaller, less visible discounting fills that gap. The same read applied to the major sale windows shows how much of that spike is incremental. And the evidence to defend a full-price push exists before someone else questions whether it is working.
A headline return metric credits every dollar of discounted revenue without subtracting the margin the discount gave away, and once it is already acknowledged that offering a discount reduces revenue on its own terms, the metric being reported may be flattering the team by design. Moving to a version of the number that nets out promotion and production cost quietly, before anyone else notices the gap, is different from having someone else discover it in the numbers first. The area holding the most confidence right now is exactly the one most exposed if the suspicion is correct.
A version of the return metric that accounts for the margin given away, calculated from existing promo and production history rather than a new reporting requirement, closes that gap. The number gets restated on its own terms and its own timeline. Whether the suspicion was right becomes known before anyone else finds out.
How a promotional or tentpole calendar gets decomposed decides how much of the client's revenue spike gets credited to the agency's work versus to the calendar itself, and the client is disputing the chosen method rather than accepting it. Holding ground on a methodology nobody on the account built, with only agency authority behind it, or changing it and explaining why the original approach was wrong, are the only two paths, and either one puts the account at risk. The client's spikes happen whether or not the agency does anything, and if the readout cannot separate the two, the account reads the relationship as the calendar doing the work.
A decomposition built on evidence from the client's own promotional history, not a fixed assumption defended by authority alone, changes that. A method the client's own analysts can inspect replaces one they can only accept or reject. The account gets kept by showing actual contribution, not by winning an argument about method.
Discount incrementality is the share of a promotion's revenue that would not have happened without the discount, as opposed to revenue from customers who would have bought at full price anyway. A promotion can report a strong return on paper while most of that revenue is the second kind. Across 568 models from 90 advertisers, a median of 15% of promotional-period revenue was attributable to the promotion itself, with the middle half between 6% and 23%. That is what the model could attribute, not proof that the rest was cannibalised.
By reading the pattern already present in existing promotional history rather than running a new controlled discount test. Periods with different discount depth, timing, or duration leave a signature in sales that a genuine incremental effect would produce, and that signature is measurable without holding back a promotion that would otherwise run.
A return-on-spend metric usually credits all revenue from a promoted period without subtracting the margin the discount gave away, which can make a promotion look successful even when much of that revenue would have happened at full price. A margin-aware version of the same metric nets out the discount cost, which changes which promotions look worth repeating.
Direct-to-consumer brands trying to grow full-price sales relative to sale-period revenue, and agencies whose promotional or tentpole decomposition is being questioned by a client. Both are asking the same underlying question from different seats: how much of this revenue spike is the calendar, and how much is genuinely incremental.
When there is not enough promotional history or variation in discount depth to separate a genuine effect from the shape of the calendar, when the decision needed is whether to approve a specific markdown right now, or when a promo mechanic is being run for the first time with no prior instance to compare against. In those cases the honest answer is a wider range, not a confident one.