A snack-food brand paid for shelf space it wasn't getting. We built an AI vision flow: field reps snap one shelf photo, the planogram check runs in seconds, and gaps get fixed before the rep leaves the aisle. Display compliance climbed from 64% to 92%.
Typical results
30 sec
Photo to compliance score
4 min
Per store check
180
Stores covered monthly
Figures reflect typical results for this kind of automation.
The challenge
The brand negotiated facings and secondary placements across 180 stores, but had no reliable picture of what shelves actually looked like. Manual checks took 25 minutes per store, produced subjective notes, and gaps surfaced weeks later in sales data. Measured compliance hovered around 64%, and every retailer conversation started from anecdotes.
Before and after
Display compliance rate
+44%Minutes per store check
−84%Days to fix a shelf gap
−93%The automation
We gave every field rep a simple flow: photograph the shelf, get a verdict. AI vision compares the photo against the store's planogram — counting facings, verifying placement and reading price tags — and returns a compliance score with specific gaps to the rep's phone in about thirty seconds. The rep fixes what's fixable in the aisle and flags the rest. Every check rolls up into a dashboard by store, chain and region.
How it works
- 01
Snap the shelf
The rep photographs the display from the aisle — one photo covers the whole planogram section.
- 02
AI reads the display
Vision AI counts facings, checks placement and reads price tags against the store's planogram.
- 03
Score in 30 seconds
A compliance score with the specific gaps returns to the rep's phone in about thirty seconds.
- 04
Fix in the aisle
The rep corrects facings and pricing immediately, or flags store-level issues to the account manager.
- 05
Roll up by chain
Every check aggregates by store, chain and region — evidence for the next trade-terms review.
Starts from
- Shelf photo from the rep
- Planogram spec
Delivers
- Instant compliance score
- Gap alerts to the rep
- Chain-level compliance dashboard
The impact
Display compliance typically rises from about 64% to 92%. Store checks take four minutes instead of 25, so reps cover more doors per day, and gaps that once lingered two weeks are fixed the same visit. The brand walks into retailer reviews with photographic proof instead of anecdotes.
Questions about this automation
What does the AI check in a shelf photo?
The vision model compares the photo with the planogram for that store. It counts facings per product, checks that each item sits in its agreed position and reads the shelf price tags. The rep gets a compliance score and a list of specific gaps, usually within about thirty seconds.
What if the photo is unclear or the AI is unsure?
If the photo is blurry, cut off or taken at a bad angle, the rep is asked to retake it before scoring. Items the model cannot identify with confidence are marked for the rep to confirm, so the score never rests on a guess. Store-level problems can be flagged to the account manager.
What is needed to set this up?
The brand needs planogram specs per store or store format, reference images of its products for recognition, and a simple phone flow for reps. Results feed a dashboard by store, chain and region. Simple automations like this are usually live within days, often starting with one chain and a few product lines.