Kroger Precision Marketing, the grocer's in-house advertising arm, posted 24% profit growth in the second quarter of 2026, according to Kroger's financial earnings release reported by Digiday. That marks the division's strongest quarterly gain since 2021 and handily outpaces the low single-digit operating margins typical of grocery retail. The mechanism: Kroger sells anonymized, first-party purchase data back to the brands on its shelves, allowing CPG advertisers to target shoppers who actually bought their competitor last week or who exhibit category intent. Brands pay for that precision.

Kroger operates Precision Marketing as a closed-loop ad platform. A brand buys display, sponsored product, or sampling campaigns inside Kroger's digital properties and physical stores, then receives attribution tied to actual basket data. The grocer does not sell raw customer files; it rents access to audience segments built from 84 million loyalty cardholders scanning items at checkout. Advertisers bid on placements, Kroger delivers the impression, and both parties see whether the exposed household converted at the register. That closed loop commands higher CPMs than open-web display because the outcome data sits inside one walled garden.

The profit jump reflects two compounding tailwinds. First, CPG brands face rising customer acquisition costs on Meta and Google, pushing media buyers toward retail media networks where purchase intent is observable and attribution is deterministic. Second, Kroger controls a scarce asset: years of sequential purchase history per household, including private-label adoption, basket size, and cross-category behavior. Competitors can buy grocery user data from data brokers, but they cannot reconstruct multi-year, SKU-level purchase sequences at Kroger's scale. That scarcity lets Kroger price its audience access above commodity programmatic rates.

A small physical-product brand with modest budget can run the same play by treating your customer file as an owned media channel and then renting it back to complementary brands. Start with your current customer list—email, SMS, or a loyalty cohort if you operate a physical location. Segment the file by purchase recency and category behavior. Approach a non-competing brand whose buyer overlaps yours and offer a co-marketing test: they pay you a flat fee or rev-share to send one dedicated email to your list, you write the copy and control send timing, and you both track conversions via unique discount codes. Cost to set up: zero new software if you already run email. A 500-person house file sending one partner email per quarter at $200 per send generates $800 annual ancillary revenue with fifteen minutes of coordination per campaign. Scale the model by recruiting three to five complementary brands—soap, candle, snack, apparel—and rotating them across your send calendar. Your customer data becomes an ad network with 100% margin because the infrastructure already exists.

The broader pattern: any brand with a defensible customer file and repeat purchase data now operates a latent media business. Kroger simply industrialized the model at grocery scale, but the arbitrage—charging more for targeted access than the cost of the audience relationship—works at 50 customers as well as 50 million.