Google Analytics — but for every shelf in the store. The first platform that measures what shoppers actually do at the shelf, in real time.
Brands spend $500B+ a year on trade marketing.
Nobody knows what works. We do.
No cameras · No Wi-Fi needed · No IT integration · No PII, ever
Trusted for 7 years across 14 markets by the world's top consumer brands and retailers.
Perfect Data has worked with Asahi, Kompania Piwowarska, Castel, CEDC, Coca-Cola HBC, Danone, Ferrero, Heineken, JTI, PepsiCo, Maspex, JUUL, Nestlé, Nutricia, Xiaomi, Rossmann, Samsung, Stock and Eurocash.
Perfect Data has been building and deploying shelf sensors since 2018 and has measured shopper behaviour in stores across 14 markets as of July 2026, including 400 smart end-caps operating in a single retail network in Poland.
The same real-time shelf data answers different questions for brands and for retailers.
You spend millions on in-store activation. We measure attention, interaction and purchase at every display — so you invest in what works and cut what doesn't.
Smart end-caps with sensors, digital signage and electronic shelf labels — sold to brands as retail media with proof of performance attached.
Modular IoT hardware that installs in minutes on existing fixtures — and streams shopper behaviour to your dashboard every second. In multiple ways, on any display.
Camera-free microwave sensing of shopper presence, paths, stops and engagement — translated into shelf-level KPIs like footfall, stop rate and dwell.
A slim, battery-powered platform that slides under any existing display and reads sell-out, hit rate and stock levels — with zero IT integration.
Smart shelves plus digital signage that reacts to shoppers in three stages — and reports exactly how content converts into pickups and sales.
Online, you'd never run a campaign without analytics. Perfect Data gives physical retail the same discipline — live reports your whole team can read in a browser, not a PDF three weeks later.
| Google Analytics online | Perfect Data in-store | |
|---|---|---|
| Impressions | → | Attentionshoppers passing & noticing the display |
| Clicks | → | Interactionsproducts picked up from the shelf |
| Bounce rate | → | Put-backspicked up but returned to the shelf |
| Conversions | → | Purchasesverified against sales data, SKU by SKU |
Retail media is the fastest-growing ad market — and most of it happens in stores, where nobody measures it. Perfect Data closes that gap: every screen and every shelf reports attention, engagement and sales, exactly like a digital campaign.
Sell end-cap screens and smart shelves to brands as measured media — with per-campaign reporting instead of promises. Proven at scale on 400 smart end-caps in one network.
Attention, pickups and sales per store, per SKU, per creative. Compare in-store campaigns with your digital dashboards — same metrics, same rigour.
Three-stage triggering: stand-by content, a shopper approaches, a product is picked up — each stage plays different content and each transition is measured.
Google Analytics made every ad on the web accountable.
We do the same for every shelf in the store.
Sensors retrofit onto your existing displays and shelves. No store network, no electrician, no retailer IT project.
Patented sensors capture attention, interaction, pickup and purchase — the full journey from pass-by to basket.
Data streams to secure cloud dashboards. Watch every display, store and campaign perform in real time.
Our analysts turn the data into recommendations: what to keep, what to move, what to stop paying for.
Every number below comes from a controlled in-store study: same SKU, same price, same promo — only the execution differs.
In a smart display program measured by Perfect Data between July and October 2024 in Poland, a premium beer brand grew sales 55% year-on-year versus 14% in matched control stores — fully incremental, with zero cannibalization of the core range.
In a six-month controlled study conducted by Perfect Data across a franchise grocery network in Poland, 400 sensor-equipped end-caps increased promotional sales by 52% versus a matched control group, reached 29% more shoppers, and kept a 5% residual uplift after the promotion ended.
In a six-month Perfect Data study for a beverage category leader across 24 supermarkets in Poland, smart displays raised shopper attention by 95%, interaction by 71% and conversion by 42% compared with a standard promotional stand and control stores.
Client names available under NDA. Ask us for the full methodology pack.
A Perfect Data checkout-zone activation in Poland lifted one category 38% and another 19%, adding over PLN 4,000 of net revenue per activation per year.
A new fixture measured by Perfect Data delivered a 29% volume uplift and 51% share of category volume — the data that justified the client's rollout budget.
In a live Perfect Data A/B test, a revolutionary planogram grew volume 27% versus 9% for the evolutionary variant — measured before any national rollout.
Dynamic screen communication measured by Perfect Data raised shopper interactions 91% versus static material and lifted conversion 15%.
Perfect Data sensors held on-shelf availability at 90%, translating into 12% more volume — the measured basis for a pay-for-performance trade agreement.
In a Perfect Data promotion evaluation, 58% of shoppers interacted with the promo display and volume grew 19% — the promotion earned its slot back.
Five packaged services built on the same platform — each one answers a question you're already asking.
Does my display actually deliver a sales uplift at the point of purchase?
dashboard + reportBuild a data-backed case for more display space before your next retailer negotiation.
negotiation packWhich displays truly deliver — and which only take up space? One score per display: the Shelf Performance Index (SPI), 0–100.
live SPI scoreboardTest 3–5 planogram scenarios in real stores and get a GO / NO-GO before rollout.
go / no-go reportContinuous measurement and quarterly recommendations to grow the whole category.
action roadmapWe measure behaviour, not people. Our sensors physically cannot capture faces, images or identities.
Microwave and weight sensing only. There is no lens anywhere in the system.
No personal data is collected — GDPR-compliant by design, not by policy.
Proprietary, patented algorithms turn sensor signals into transactions.
Enterprise-grade Microsoft cloud. Member of Microsoft "Store of the Future".
Two decades on the client side of global beverage and FMCG brands — and an engineering team that has been building shelf sensors since 2018.
21 years of managerial experience within commercial function, mainly in sales and marketing of international companies — Heineken, Red Bull, Coca-Cola Hellenic as well as EMPIK Group and POSM Factory. Vast experience in creating, developing and commercialization of startups including POSM Factory and Perfect Data.
21 years of managerial experience in sales, marketing, financial controlling and audit as well as technical service in Coca-Cola Hellenic. Vast experience in commercialization of start-ups and real estate business as well as capital investment activity.
22 years of managerial experience in sales and business development functions of companies from IT, building construction, and steel processing industries, including many years of management and optimizing of technical & production functions of MAGO S.A. Promotor of management systems based on QRM and TOC principles.
Your first call — designs pilots and research programs with brand and retail teams across Europe and the U.S.
Data analyst specialising in Power BI, data modeling and BI reporting — turns shelf-sensor telemetry into the dashboards and study reports our clients work with.
Full-stack engineer of the Perfect Data platform since 2018 — Angular front-end, .NET back-end, SQL Server, Azure DevOps, Docker and Terraform.
Developed the firmware and electronics of the IQ sensor family (2022–2024) — the embedded layer that streams shelf events to the cloud.
Designs and tests the sensor electronics behind the IQ device family — from PCB design to production batches.
Perfect Data uses microwave presence sensors and weight sensors built into the shelf. The microwave sensor detects that a shopper is present, stops and dwells; the weight sensor detects that a product was picked up, put back or removed for purchase. Neither sensor has a lens, so no image, face or identity is ever captured. Uplift is calculated by comparing test stores with a matched control group running the same SKU, price and promotion.
The Shelf Performance Index (SPI) is a 0–100 score developed by Perfect Data that summarises how effectively a single display converts passing shoppers into purchases. It combines attention rate, interaction rate, conversion rate, lost-sales index and display share, all measured by in-shelf sensors. Scores above 70 indicate a high-performing display; below 40 usually signals wasted space.
By measuring the full journey at the display — attention, interaction, put-backs and purchases — and comparing test stores with a matched control group. Perfect Data reports incremental sales per display, per SKU and per store, so trade teams can show exactly which displays pay back and which do not.
Perfect Data's system collects no personal data. The sensors measure anonymous physical events — presence, dwell time, weight change — and cannot capture faces, images or identifiers. Because no personal data is processed, the system is GDPR-compliant by design rather than by consent management.
Camera systems recognise products and people from images, which creates privacy obligations and usually requires retailer approval. Sensor-based systems like Perfect Data's use microwave and weight signals: they record that a shopper stopped or a product left the shelf, never what anyone looks like. That makes deployment faster and retailer acceptance far easier.
Minutes per display. IQ Base slides under an existing display, IQ Traffic mounts on a shelf edge, and both run on batteries with independent GSM connectivity — no electrician, no store Wi-Fi and no retailer IT project.
No. Every Perfect Data device has independent GSM connectivity and offline buffering, and streams directly to Microsoft Azure. A pilot can run end-to-end without touching the retailer's infrastructure.
Perfect Data runs three to five planogram scenarios in matched groups of real stores and measures shopper behaviour, conversion and sales for each. After the test, the brand gets a GO / NO-GO recommendation before committing to a rollout.
In-store retail media is advertising space inside physical stores — screens, smart shelves, end-caps — sold by retailers to brands. Perfect Data measures it like a digital campaign: attention as impressions, product pickups as clicks and purchases as conversions, per store, per SKU and per creative.
A standard pilot runs 90 days with one brand, one retailer and one business question, and requires no retailer IT integration. Pricing depends on the number of devices and stores — contact info@perfectdata.ai for a quote.
Only anonymous events: someone passed, stopped, dwelled, picked a product up, put it back or bought it. No images, no faces, no device IDs, no loyalty data. Individual shoppers cannot be identified or re-identified.
Trax and RetailNext are strong image-recognition platforms, typically used for planogram compliance audits and store traffic. Perfect Data takes a different approach: camera-free microwave and weight sensors that measure behaviour and sales continuously in real time, with zero personal data. For GDPR-sensitive European retailers and always-on display measurement, that structural difference is often decisive.
The terms we use across studies and dashboards — defined the way we measure them.
Book a 30-minute demo. We'll show you live data from real stores — and how a pilot would work for your category.
Book a demoOr start small: 90-day pilot · 1 brand × 1 retailer × 1 question · no retailer IT integration
Fresh from the store floor.
Films, case stories and behind-the-scenes from live deployments — straight from our channels.