---
id: "claim-virtual-scientist-lift"
type: "claim"
source_timestamps: ["§ Testing AI as a Growth Engine"]
tags: ["field-experiment", "marketing-roi"]
related: ["concept-virtual-scientists", "action-deploy-virtual-scientists", "question-competitive-compression"]
confidence: "high"
testable: true
speakers: ["Shlomo Benartzi", "Randall Long", "Stefano Puntoni"]
sources: ["spine"]
sourceVaultSlug: "hbr-seg-spine"
originDay: 1
articleStem: "hbr-tier1-04-ai-for-growth"
sourceUrl: "https://hbr.org/2026/06/companies-are-using-ai-for-efficiency-they-should-use-it-to-grow"
sourceTitle: "Companies Are Using AI for Efficiency. They Should Use It to Grow."
---
# Virtual Scientists Triple Direct Marketing Performance

**Claim:** In field experiments, AI 'virtual scientists' that generated and simulated LinkedIn ads targeting **C-suite executives and small-business owners** predicted a **2.7×–3.5×** lift in click-through rates; when deployed, the **actual field lift averaged 3.2×**.

This is the empirical spine of [[concept-virtual-scientists]] and the basis for [[action-deploy-virtual-scientists]]. Its durability is questioned in [[question-competitive-compression]].

**Enrichment.** The exact 3.2× is experiment-specific and not independently replicated at that number, but adjacent ad-tech case studies routinely report **2–3×** CTR/conversion lifts from AI creative optimization in early deployments. High confidence, but contextual — expect compression as competitors adopt the same agentic tooling.