- index
say hellohello@simplestuff.ai
elsewherelinkedin · notes
simple stuff · 2026
case studies / online grocery · dach

the promo plan that picks itself.

how a major european grocer took promo margin from 17% to 27% across four DACH fulfilment centres - by picking smarter, not cutting deeper.

sectoronline grocery
regionDACH
footprint4 fulfilment centres
what we didmap · build · embed
grocery fulfilment centre
promo operations · online grocery fig. 01
the outcome
+10pp promo margin lifted, 17% → 27%
+€2.6M added annualised promo margin
promo revenue in the 5‑week window (+275%)
~32% total margin held - discount depth slightly down
the challenge

promos picked on gut feel, priced to clear.

across four dach fulfilment centres, promo nominations were made every week - and largely on gut feel.

the demand forecast over‑estimated promo uplift by 9–14× on average, so discount depth became the only reliable lever: cut deep enough and the stock clears.

promo share of net revenue sat at 7.6%, and margin quietly diluted by discounting that wasn't aimed at anything in particular.

what we did

a system, not a dashboard.

01

map

we traced how nominations actually got made across the four fcs - the data buyers leaned on, the data they ignored, and where the 9–14× forecast error came from. the lever was always depth; the full constraint picture was never in the room.

  • nomination tracing
  • forecast‑error analysis
  • constraint mapping
02

build

an agent that pulls campaigns, supplier agreements, sales history and pricing from the data warehouse every day - then builds the optimal promo portfolio per fc, balancing margin, depth, supplier mix and inventory, and writes it back into the promo system.

  • portfolio agent
  • warehouse integration
  • write‑back to promo system
03

embed

same buyers, same suppliers - different prep. the agent does the constraint math every cycle; the team makes the calls and tunes the objectives, now with the whole picture in front of them.

  • buyer enablement
  • objective tuning
  • ongoing support
before → after · promo margin
before
17%
promo margin. nominations on gut feel, a forecast off by 9–14×, and discount depth as the only lever - promo share stuck at 7.6%.
after
27%
promo margin. the agent rebuilds the optimal portfolio per fc every day - margin, depth, supplier mix and inventory weighed together. revenue tripled; total margin held.
the takeaway
we didn't discount harder. we discounted smarter - and tripled promo revenue while margin held.
- measured: iso wks 9–13 · 2025 vs 2026
what changed

tripled promo revenue. margin held.

over a five‑week window - iso weeks 9 to 13 - promo margin moved from 17% to 27%. ten points, worth about €2.6m annualised.

promo revenue tripled in the same window: +275%, +€6.43m. total margin held at roughly 32% - and discount depth was slightly lower than before.

the lift didn't come from cutting deeper. it came from picking better - the agent weighing the full constraint set every cycle, where a person could only ever see the headline campaigns.

your promo plan, or whatever's run on gut feel

there's a version of this for your ops.

hello@simplestuff.ai →

not ready to talk?

get the newsletter. the same thinking and frameworks we use with clients - what's working, what isn't, and how to tell. free.