Beyond the Basket: The Rise of AI Consumption Intelligence
Retail media 1.0 told you what shoppers bought. Retail media 2.0 predicts when they'll run out — and intercepts them at the exact moment of replenishment. This is the moat no grocery RMN has built. And it only exists in convenience.
Impression
Purchase
Consumption
Return (DRS)
Retarget
Every retail media network in the UK targets based on what shoppers bought. None of them know when they'll run out.
The entire edifice of retail media targeting is built on a single data layer: purchase history. Tesco Media knows you bought Gillette Fusion razors six weeks ago. Sainsbury's Nectar knows you bought Olay moisturiser last month. These signals are valuable — they are deterministic, real, and privacy-compliant in a way that third-party cookie data never was. But they are also, fundamentally, retrospective.
Retrospective targeting creates a structural inefficiency that every brand manager intuitively understands: the shopper who bought shampoo three weeks ago is probably fine. The shopper who bought shampoo three weeks ago and uses it daily is running out right now. These two shoppers look identical in a purchase history dataset. But they have completely different media receptivity profiles. One is a low-intent background audience. The other is a high-intent active replenishment seeker.
The current generation of retail media cannot distinguish between them. Corner Collective's AI Consumption Intelligence engine can. By learning individual replenishment cycles across categories, SKUs, and household compositions, our system predicts the depletion window for every shopper in the network — and activates targeting sequences timed to the moment of need, not the moment of purchase.
Purchase data tells you what a shopper wanted at a point in time. Consumption intelligence tells you what they need right now. The gap between those two signals is where most retail media budget is being wasted — targeting audiences with messages they are not yet ready to receive.
Retail media has evolved through three generations. Most UK brands are still on generation one.
To understand why AI consumption intelligence represents a competitive advantage and not just a feature upgrade, it is necessary to understand the generational evolution of retail media targeting — and where the industry is positioned today.
Keyword-bidding on retailer search engines. High-intent but limited to active search sessions. No audience building, no offsite extension, no measurement beyond last-click.
Loyalty card data enables offsite programmatic targeting. Closed-loop attribution measures in-store lift. Dominant model at Tesco, Sainsbury's, Ocado today. Retrospective — based on past purchase.
Predictive depletion targeting. System learns individual replenishment cycles. Activates at the moment of need. DRS confirms actual consumption, closing the loop entirely. Uniquely available in convenience via Corner Collective.
The generational transition matters because each generation has delivered exponentially higher ROAS than the previous one. The move from keyword bidding to first-party audience targeting was worth an average 3-5x improvement in conversion rates across early adopters. The move from retrospective first-party data to predictive consumption intelligence represents an equivalent or larger step change — because it solves the single biggest inefficiency in current retail media: timing mismatch.
The Deposit Return Scheme is a compliance event for most retailers. For Corner Collective, it is the most powerful consumption signal in FMCG history.
England's Deposit Return Scheme (DRS) is mandated to launch in October 2027, joining existing DRS programmes in Scotland, Malta, Romania, Poland, and Austria. The scheme requires consumers to pay a deposit on eligible beverage containers — typically carbonated soft drinks, water, beer, and cider — and return them to a registered return point to reclaim the deposit.
Every major convenience and forecourt retailer will become a return point. Every return transaction will confirm, at SKU level, that a specific product was purchased, opened, and consumed. For the first time in the history of retail data, a system will exist that tracks not just purchase intent but confirmed consumption — the product was bought, the liquid was drunk, the container was returned.
Ad Impression
Served via TTD, DV360, Yahoo DSP
In-Store Purchase
Confirmed via ePOS / loyalty
DRS Return
Consumption confirmed at SKU level — unique to Corner Collective
AI Retargeting
Predictive depletion — serve next offer at exact moment of need
The strategic implications for FMCG brands are profound. A brand running a campaign on Coca-Cola Zero Cherry can, for the first time, know that a specific shopper bought the product, consumed it, and is likely to run out in 3-4 days. The next ad impression can be served precisely at the replenishment window — not randomly during the week following purchase. The difference in conversion rates between a well-timed replenishment nudge and a randomly-timed retarget is measured in multiples, not percentages.
Consumption patterns change with life stage. A first-time parent buying nappies has a predictable and recurring consumption cycle. A student buying energy drinks has a different one. Corner Collective's AI learns these individual cycles across the entire network — building what is effectively a consumption biography for every shopper. As life stage changes, the model adapts. The targeting gets more precise over time, not less.
What AI consumption intelligence means for your media plan: A direct comparison.
The practical implication for FMCG media teams is not abstract. Every brand running retail media today is choosing between targeting approaches that have measurable differences in conversion efficiency. The table below maps the current landscape across three approaches: standard grocery RMN targeting, enhanced first-party targeting, and Corner Collective's AI consumption intelligence.
The CPM advantage in convenience RMNs is a structural feature of market development, not a permanent discount. Early adopters of Tesco Media and Ocado Advertising benefited from significantly lower CPMs in the 2021-2023 period before the market normalised. The same dynamic applies to the UK convenience RMN market today — early-entering brands will establish audience relationships and attribution models at pre-competitive pricing, building a performance dataset that later entrants will take years to replicate.
The ROI case for AI consumption intelligence. Quantified.
The commercial case for adopting AI consumption intelligence as a media capability rests on three quantifiable improvements over standard retail media targeting. Each is independently significant; combined, they represent a step change in FMCG media efficiency.
Standard retail media reaches audiences at random points in their consumption cycle. Consumption intelligence targets the final 20-30% of the depletion window — when replenishment intent is highest. Industry data suggests this timing precision alone improves conversion rates by 20-30% (McKinsey) when applied to personalised offers. The same ad, served at the right moment, converts at a materially higher rate than the same ad served at the wrong moment.
Standard retail media attributes approximately 70% of actual converted sales — the remainder are lost to channel fragmentation, purchase delays, or cross-channel attribution failures. DRS-enabled consumption attribution adds the third confirmation: the product was bought AND consumed. This completes the attribution chain and gives brands accurate LTV data for convenience channel shoppers for the first time. Correct attribution data leads to correct budget allocation — which compounds over planning cycles.
Because the convenience channel is currently invisible in FMCG media models, any sales volume driven through Corner Collective's network is, by definition, incremental — it cannot have been attributed to existing media spend. For categories where convenience accounts for 25-40% of volume, this represents a measurable and previously uncaptured contribution to total brand performance. iROAS for well-executed closed-loop convenience campaigns is benchmarked at £25-35 — consistent with the best-performing grocery RMN campaigns.
Depletion Prediction
Learns individual replenishment cycles. Predicts exact depletion window. Activates media at the moment of need — not retrospectively.
DRS Consumption Loop
The only UK RMN integrating DRS return data. Confirms consumption at SKU level. Closes the attribution loop from impression to consumption and beyond.
CDP + DSP + Attribution
Resolves identities across 50,000+ c-stores. Activates via TTD, DV360, Yahoo DSP. Measures iROAS using the same methodology as Tesco Media.
Active Partnerships
Shell GO+, Bestway, Shopmate ePOS (4,000+ stores), GreenJinn (1.5M shoppers). LOIs: Zenith, Publicis, independent buyer.
The only UK retail media platform with true consumption data.
We built the infrastructure that grocery RMNs cannot replicate: a closed-loop AI system that learns when shoppers run out of products, confirms consumption through DRS returns, and activates targeting at the exact moment of need. Patent pending.
Get a Platform BriefingWant to Learn More?
Get in touch with our team to discuss how Corner Collective can transform your retail media strategy.