A Warehouse Club Just Told Advertisers It Can Identify 800,000 Households That Will Buy a Specific Product Within Twelve Months, With 94% Accuracy
Sam's Club Connect introduced "Predictive Precision Targeting" on September 17, 2026, positioning its retail media network as something more than a place to reach past buyers. The pitch: stop paying to remind people who already bought your product and start paying to find people who haven't bought it yet but will.
The evidence comes from a P&G pilot. Sam's Club used its membership data and more than 400 intent signals to target households with no prior purchase history in the Cascade Platinum Plus category. The model predicted which members would buy and how much they would spend. After twelve months, actual purchases matched the forecast at 94% accuracy, and the targeted group spent 71% more than their initial purchase during the campaign window.
The Membership Advantage, Quantified
Retail media networks have always claimed first-party data superiority over social and search platforms. But most of that superiority has been backward-looking: we know what people bought, so we can find more people like them. Sam's Club is making a different argument.
The new targeting products include a "Predictive Prospect Targeting & Lifetime Value" score that identifies high-potential prospects through focused campaigns built around specific objectives. The system watches for purchases of soundbars and gaming consoles to find people who are about to buy a television. It looks at long-term purchase history across related product categories to identify behaviors like the desire to upgrade products or early signs of intent to change brand preferences.
This is where the membership model creates structural advantage. As Lex Josephs, then VP of ad sales and general manager, explained when Sam's Club launched its Member Access Platform in 2022:
Lex Josephs
Because it's 100% traceable tender where everything is related back to a credit card, you know everything about a member.
Every transaction, online or in-club, ties back to a known household. That longitudinal view is what makes predictive modeling possible at this scale.
What the CFO Needs to See
The P&G pilot delivered a 2.3x increase in marketing outcome, according to MediaPost's coverage. But the more interesting number is the 94% accuracy on a twelve-month prediction. If that holds across categories and advertisers, it changes how you model CAC payback.
Most retail media attribution is still reactive. You run a campaign, measure lift against a control, and report ROAS. The problem is that ROAS on past-buyer targeting often overstates incrementality because those buyers might have purchased anyway. Predictive targeting on non-buyers is a cleaner test of media impact.
Sam's Club Connect also introduced "Omni-Impact" technology that measures attribution and media-mix modeling across in-club demos, direct mail, Scan & Go, and Meta Platforms products. The tool launched in mid-2025 and now includes AI-assisted recommendations on where to shift budget. This is the closed-loop measurement that makes retail media defensible in a board presentation: you can trace the ad impression to the in-club purchase, not just the online conversion.
Sam's Club Connect's measurement page claims it is "possibly the only retail media platform to connect Search and Sponsored Product Ads to offline sales." That's a bold statement, but the membership model makes it plausible. When every transaction requires a membership scan, you don't need probabilistic matching to connect digital exposure to physical purchase.
The Assumptions Worth Stress-Testing
Before reallocating budget, run the sensitivity analysis. The P&G pilot is one advertiser, one category, one campaign. The 94% accuracy figure needs context: what was the baseline purchase rate in the control group? What was the cost per household reached? How does the model perform in categories with longer consideration cycles or lower purchase frequency?
The predictive models rely on AI and machine learning trained on Sam's Club's membership data. As Harvey Ma, VP and general manager of Sam's Club Connect, noted at Cannes Lions, the membership model gives them visibility into virtually every transaction. But model performance degrades when you move outside the training distribution. A model that predicts dishwasher pod purchases may not transfer cleanly to consumer electronics or apparel.
The other assumption to test: incrementality. Sam's Club uses third-party partner Circana to validate incrementality, ROAS, and sales lift. That's better than self-reported metrics, but incrementality measurement in retail media still has methodological debates. Ask for the holdout design, the match rate between exposed and control groups, and the confidence intervals on lift estimates.
A Pilot Design for the Skeptical
If you're considering testing Sam's Club Connect's predictive targeting, here's a two-week pilot structure:
- Select a product with measurable purchase frequency and a clear non-buyer segment.
- Request the Predictive Prospect Targeting audience for households with no prior purchase history in your category.
- Run a matched holdout at 10-15% of the target audience.
- Measure conversion rate, cost per acquisition, and incremental ROAS against the holdout.
- Track the cohort for 90 days post-campaign to validate the lifetime value predictions.
The risk is that predictive targeting costs more per impression than past-buyer retargeting, and the payback period is longer. The opportunity is that you're reaching genuinely incremental buyers, not subsidizing purchases that would have happened anyway.
Where This Fits in the Retail Media Stack
Sam's Club rebranded its Member Access Platform to Sam's Club Connect in August 2026, signaling a shift from transactional ad network to what the company calls a "Retail Experience Network." The platform has been building toward this for years, adding in-club attribution for search and sponsored product ads in 2023 and expanding to connected TV, video search, and 3D interactive product pages by late 2023.
The competitive context matters. Amazon's retail media network dominates on scale. Walmart Connect has the store footprint. Sam's Club is betting that membership density and predictive precision can command a premium even with a smaller audience. For advertisers, the question is whether that precision translates to lower effective CAC and higher customer lifetime value.
The P&G pilot suggests it might. But one pilot is not a strategy. Model the assumptions, run the holdout, and let the data tell you whether predicting future buyers is worth more than retargeting past ones.
