Margin & Markdown Sentinel · Brief 02
Q3 margin inversion in Beauty — who is funding the discount
Beauty · Q3 FY25 (Aug–Oct 2025) · markdown rate by banner → price realisation → where the premium went
The one line you configure
Beauty markdown is up on the quarter. Is that a promotional decision we took, or is it the shape of what the vendors are refusing to fund?
In a nutshell
It is the vendors. National-brand beauty markdown rate compressed to ~1.0% in Q3 against ~3.0% off-peak — they held price and funded no clearance — while owned-brand beauty ran ~7.9% to keep share of cart in the same audience. The gap is 6.9 points against an expected seasonal gap of 1 to 2. Two signals confirm where the money went: national-brand price realisation is ~8% above its own trajectory, and marketplace 3P take rate on beauty sits ~8% below the GMV it should be riding. The premium is parked in first-party vendor margin.
At a glance
What moved
What · the two banners swapped places inside one quarter
Off-peak the two banners sit within two-tenths of a point of each other. In Q3 one falls to a third of its normal rate and the other more than doubles — the same shelf, the same audience, opposite directions.
Where the price premium ended up · three datasets, one quarter
| Signal | Q3 FY25 | Its own baseline | What it says |
|---|---|---|---|
| National_Brand price realisation (T1) | +8% | on trajectory | Vendors are not giving promo support |
| Retail-media iROAS, mass + upper-mid tier (T3) | +45% | baseline | Brand equity is winning the auction at full price |
| Marketplace 3P beauty take rate (T2) | −8% | GMV trajectory | The premium is not sitting in 3P take either |
| Owned_Brand markdown rate (T1) | 7.9% | ~3.3% | First-party margin is paying for the gap |
Price realisation holds, ad efficiency rises, 3P take falls, own-brand discounting doubles. Every one of those is consistent with vendor pricing power and none of them is consistent with a merchandising or marketing failure.
Why · the gap is one department and, inside it, two category families
Contribution analysis of the markdown-rate gap. It is not a book-wide vendor posture — it is skincare and cosmetics, in the two regions with the highest beauty mix, against the audience both banners are bidding for.
Why the quarterly read is the one that catches it
Month by month each banner looks ordinary — national brand simply is not discounting, owned brand simply is. Only the two read against each other, over the quarter the contracts are written in, shows a 6.9-point gap where 1 to 2 is normal.
So what
This is a contract conversation, not a pricing one. Owned-brand beauty is spending 4.6 points of markdown rate a quarter to stand next to vendors who fund none, and both the ad auction and the 3P take rate confirm the premium is sitting in first-party vendor margin. Q4 renewals are the lever; the categories with the widest gap are the list to open with.
The points that matter
The gap is the finding, not the level
Either banner's markdown rate alone reads as normal seasonal behaviour. Read against each other, 6.9 points where 1 to 2 is expected is a pricing-power signal.
6.9ppvs 1–2ppOwn brand is paying for it
Owned-brand beauty more than doubled its markdown rate in the quarter while national brand cut its own to a third — the discount that keeps the shelf competitive is coming entirely out of first-party margin.
7.9%vs 3.3% normalThe ad auction says the same thing
Mass and upper-mid advertisers — the national-brand vendors — are running 45% above baseline iROAS. They can hold price and still win the placement.
+45%iROASRecommendation
Open Q4 contract renegotiations with the top national-brand beauty vendors on the Skincare and Cosmetics categories, and reweight retail-media sponsorship so some of the margin comes back as co-op spend rather than as owned-brand markdown.
Questions it already answers
Is this a promo decision we made?
No. National-brand markdown fell to ~1.0% on its own; owned-brand rose to ~7.9% to keep pace on the same shelf. The decision being reflected is the vendors' refusal to fund clearance, not a campaign we ran.
Where did the price premium go?
Into first-party vendor margin. Price realisation on national brand is ~8% above its trajectory, and marketplace 3P take rate on beauty is ~8% below the GMV it should track — so it is not in 3P take either.
Which categories do we open the renegotiation on?
Skincare and Cosmetics — 58% of the gap between them — concentrated in the Northeast and West, which carry 51% of the beauty mix.