63 Cents of Every New Marketing Dollar for E-Commerce Is Going to AI and Cashback Apps — Not Ads

New Northwestern University Retail Analytics Council and Minty research indicates e-commerce marketers plan to allocate 63 cents of every new marketing dollar toward AI commerce tools, as consumers increasingly use AI to identify cashback, lower prices, promotions, and alternatives. The shift signals a meaningful reallocation of e-commerce advertising budgets away from traditional digital-advertising channels, benefiting AI-enabled commerce and marketing platforms while posing a competitive risk to incumbent ad channels.

Analysis

The investable issue is not aggregate ad spend, but attribution ownership. If shopping discovery shifts from keyword/search and social-feed browsing toward AI-mediated comparison, merchants will favor channels that can prove incremental conversion and closed-loop ROAS. That favors retail-media ecosystems with first-party transaction data—AMZN, WMT, TGT and SHOP—over open-web intermediaries such as TTD and lower-intent social inventory such as SNAP and PINS, where measurement is less defensible when the consumer arrives having already narrowed the purchase set.

Near term, this is more likely a budget-reallocation narrative than a material revenue event: survey-based marketer intentions are not evidence of executed spend or durable CAC savings. The 1-3 month catalyst is Q3/Q4 commentary on advertiser demand for AI campaign tools, conversion pricing, and retail-media take rates; the 6-18 month consequence is multiple divergence based on whether platforms retain the consumer decision layer or become commoditized fulfillment/ad inventory. The contrarian view is that AI shopping agents may increase total performance-marketing spend by improving conversion rates, benefiting META and GOOGL if their AI targeting and shopping surfaces remain the final measurable touchpoint rather than disintermediating them.

The key falsifier is not AI-product adoption headlines but reported ad-price and conversion trends: sustained deceleration in SNAP/PINS revenue per impression or TTD platform spend alongside accelerating AMZN/WMT ad growth would validate the substitution thesis. Conversely, stable social/search pricing, improving conversion metrics, and no deterioration in merchant CAC would indicate AI is additive demand generation rather than a channel disruptor.

Key Decisions for Investors

• Watch, do not immediately trade, the survey signal. Require Q3 advertiser commentary and reported retail-media growth to confirm that budget intent has converted into spend before sizing directional exposure.
• On confirmation, express a 3-6 month relative-value trade: long AMZN and/or WMT retail-media exposure versus short SNAP or PINS. The thesis is superior first-party purchase data and direct conversion measurement; exit if SNAP/PINS revenue growth and pricing remain resilient relative to retail-media peers through the next earnings cycle.
• Avoid a blanket short of META or GOOGL. Both can retain economics if AI raises conversion and their automated buying products preserve closed-loop measurement; initiate only if management signals shopping-query or lower-funnel ad-share loss, not merely higher AI investment.
• Place TTD on a downside watch for 6-12 months rather than a standalone short: deterioration in platform spend growth or agency commentary on reduced open-web prospecting would make it a cleaner disintermediation expression. Offsetting risk is that AI-driven optimization increases programmatic demand rather than redirects it.
• For a lower-beta structural beneficiary basket, consider SHOP, AMZN and WMT on pullbacks over 6-18 months; their merchant, transaction, and fulfillment data can become more valuable as merchants demand auditable AI-driven customer acquisition.