CVS Claims 16% Reach Boost From AI. The Catch? Self-Measured Numbers.

CVS's retail media AI platform claims outsized gains. The numbers come from a sponsored panel, and the industry's measurement problem has not gone away.

CVS Media Exchange made headlines this week when its VP and General Manager Parbinder Dhariwal revealed performance figures for CorIQ, the pharmacy chain's AI-driven marketing platform: 16% greater reach efficiency, a 14% increase in attributed sales, and a 15% improvement in incremental ROAS for brand partners. Consumer health company Kenvue, via director Kelly Shu, claimed more than 30% sales lift alongside improvements in buyer conversion and basket size.

Compelling. Also worth noting: the panel where these figures were disclosed was co-hosted and co-sponsored by CVS Media Exchange itself, published through Adweek as a branded partnership post.

The structural conflict at retail media's core

Retail media networks occupy an unusual position in the advertising ecosystem: they sell the ad inventory and measure its performance. That structural conflict sits at the heart of what AdExchanger has called retail media's "trust problem", a problem the industry has not resolved, regardless of how the metrics are packaged.

The CorIQ figures were presented without citation to independent third-party verification. No geo-holdout test. No randomized control trial. No MRC-audited methodology. The numbers may well be accurate. They are, as stated, vendor-claimed.

Industry-wide measurement gaps undermine AI ROI claims

The measurement credibility issue is not unique to CVS. It is systemic. According to Skai's 2025 State of AI in Retail Media report, 44% of marketers report active concerns about the accuracy of retail media results, 43% say they lack clear methodologies, and 41% cannot reliably distinguish ad-driven sales from conversions that would have happened anyway. The share of marketers who say they can prove AI ROI has dropped from 49% to 41% overall, and from 54% to just 38% in retail specifically.

The ROAS metric itself is structurally vulnerable. One incrementality audit by W3Solved found a US$4.50 traditional ROAS collapsed to US$3.15 once organic conversions were factored out, a 30% overstatement. CVS's claimed "15% incremental ROAS" is notable for using the word "incremental," but without disclosed methodology, the number is unverifiable.

Dhariwal has previously acknowledged the industry's credibility gap, stating that "clear methodologies and disclosed definitions are foundational if brands and retailers want to make confident, long-term investment decisions." CorIQ's public debut does not yet meet that standard.

What CVS's data assets actually bring to the table

CVS does bring genuine data assets. Its ExtraCare loyalty program spans 90 million addressable members with a claimed 90%+ one-to-one transaction match rate, rare scale for a pharmacy-adjacent retail media network. CorIQ applies AI across media planning, real-time audience optimization (removing converted shoppers, replenishing segments), and measurement. That architecture is directionally sound.

eMarketer named AI-driven measurement a top commerce media prediction for 2026, framing it as the year "experimentation meets accountability." CVS is positioning CorIQ inside that narrative.

Why APAC brand marketers should demand independent audits

For marketing and communications leaders across Asia, the CVS announcement is a signal, not a template. APAC is already the world's largest retail media region, estimated at US$97.8 billion in 2025 with 13.5% growth, driven by Shopee, Lazada, Tmall, Flipkart, and Rakuten. The measurement opacity problem is identical, and in some cases worse. Amazon, Walmart, and other major networks each use incompatible attribution windows and methodologies, making cross-network ROAS comparison effectively impossible.

Shopee Malaysia's shoppable livestreams now drive up to 35% of weekly retail media transactions, but those numbers also come from platform-reported data, not independent audit. Digiday Research found that advertiser priority for cross-platform measurement jumped to 72% in 2026, up from 64%.

AI in retail media can improve campaign efficiency. But the measurement layer remains self-reported. For APAC brand marketers evaluating Shopee Ads, Lazada Sponsored Solutions, or regional equivalents, the critical ask is not better ROAS. It is independent incrementality testing. Until retail media networks open their measurement to external audit, the most important metric is the one they do not publish.