---
id: "claim-traditional-marketing-fails"
type: "claim"
source_timestamps: ["¶2", "§ What We Found"]
tags: ["marketing-effectiveness", "ai-behavior"]
related: ["concept-human-centric-persuasion", "claim-ratings-and-price-are-universal", "contrarian-conversion-rate-divergence"]
confidence: "high"
testable: true
speakers: ["Jafar Sabbah", "Oguz A. Acar"]
sources: ["geo"]
sourceVaultSlug: "hbr-seg-geo"
originDay: 3
articleStem: "hbr-tier2-06-ai-shopping-agents"
sourceUrl: "https://hbr.org/2026/05/research-traditional-marketing-doesnt-work-on-ai-shopping-agents"
sourceTitle: "Research: Traditional Marketing Doesn’t Work on AI Shopping Agents"
---
# Traditional marketing cues do not reliably influence AI agents

**Claim (confidence: high · testable: true):** Persuasion tactics refined over decades for human cognition — scarcity, countdown timers, strike-through pricing, bundling — do **not** work the same way on [[concept-ai-shopping-agents|AI agents]].

**Evidence:** In a simulation of **16,000 choice situations** across **4 AI models** and **8 promotional badges**, these well-known tactics showed **no stable pattern**. Depending on the model and product category, a cue sometimes increased selection, sometimes had no effect, and sometimes actively **reduced** it. This is the empirical core of the source's thesis and the failure mode described in [[concept-human-centric-persuasion|human-centric persuasion tactics]].

The corollary is that a scarce, positive set of signals *does* work reliably (see [[claim-ratings-and-price-are-universal]]), and that some "proven" CRO tactics can actively de-optimize agent conversion (see [[contrarian-conversion-rate-divergence]]).

**Enrichment / external corroboration:** Secondary coverage confirms the design — 8 promotional mechanisms (scarcity indicators, countdown timers, strike-through pricing, vouchers), 4 models ([[entity-gpt-4-1-mini|GPT-4.1-mini]], [[entity-gpt-5|GPT-5]], [[entity-gemini-2-5-pro|Gemini 2.5 Pro]], [[entity-gemini-2-5-flash-lite|Gemini 2.5 Flash Lite]]), 16,000+ rounds — and reports these tactics "often failed or backfired." The ACES framework independently finds strong model dependence and presentation biases that differ widely across agents, i.e., **no universal, stable effect** of any single promotional pattern.

**Assessment:** Well supported by both the study and independent agent-behavior research.

**Related:** [[concept-human-centric-persuasion]] · [[claim-ratings-and-price-are-universal]] · [[contrarian-conversion-rate-divergence]]


## Related across articles
- [[claim-persuasion-science-gap]]
- [[claim-sponsored-penalty]]
- [[claim-ai-ignores-implicit-cues]]