In-Store Retail Screen Buying with Basket-Level Readout

Sales effect from in-store screens is measurable, but only on a matched-store design with basket data behind it, and that design has a floor: about 60 stores per cell across six trading weeks before an effect of the size these screens actually produce, low single digits on the featured line, separates from weekly noise. Below roughly 40 stores per cell we will run the campaign and report delivery, and we will not call the result sales lift. That sentence has cost us two briefs and it still opens the conversation. Where the screens have to be: Ahead of the decision, in the aisle or on the approach to it. Entrance gantries build recall, category screens move units, and the two are not interchangeable. A panel positioned where the shopper has already reached for the pack is inventory that reports impressions and changes nothing. We map offered panel positions against the planogram before agreeing a price, and in most estates around a fifth of the offered panels come out of the plan at that step. Estates we transact: Grocery, pharmacy and specialist chains in the Czech Republic and their equivalents abroad. Domestically that is Albert, Billa, Globus, Kaufland, Dr. Max, Benu and Pilulka. Internationally it runs through retail media networks and their programmatic supply where the retailer has opened it, including Tesco, Carrefour Links, Walmart Connect in-store and Kroger, plus screen operators such as Grocery TV routed through Vistar or Broadsign. Measurement: Test and control stores matched on 52 weeks of pre-period sales, footfall band, format and region, with a synthetic control fitted on the pre-period rather than a plain average of the control group. Category and basket data from NielsenIQ or Circana where you subscribe, retailer loyalty data where the retailer will share it under contract. We read movement on the featured line and on the category, because a screen that moves your SKU by taking volume from the pack beside it is a fact your category manager needs before your CFO hears the first number. On cameras: Audience measurement units from Quividi, AdMobilize and similar vendors sit in some estates. We accept anonymous counting and dwell. We will not build a plan on inferred demographics or emotion detection, and where a retailer runs that we want the DPIA and the retention policy in writing before we place a euro. In a pharmacy environment we do not use it at all. Not included: Trade and shopper negotiation, listing fees, planogram influence, coupon and loyalty offer mechanics, packaging, in-store audio and sampling. We are not your retail sales team and we do not sit in the joint business plan meeting. Who this is not for: Brands distributed in fewer than 40 stores. Brands whose product is not on shelf in the estate carrying the screens. And anyone who wants the buy priced by store count, which is a supply metric that says nothing about how many people pass that panel holding a basket. Fees and media: Package prices are our fee. Retailer media, data subscriptions and any measurement licence are invoiced separately at cost. A readable test needs six to eight weeks in market plus a four-week post-period before the numbers settle, and we will not compress that to fit a quarter-end review.