Does Adobe's AI referral traffic growth statistic hold up under scrutiny?

Andrew Lipsman, writing on The AI Ad Economy, argues that the statistics fueling "agentic commerce" hype do not hold up once checked against independent data. His central example is Adobe's holiday-season report that "Generative AI tools drove a 693.4% increase in traffic to retail sites." Lipsman notes that Adobe's PR firm pitched the growth rate repeatedly but never supplied the underlying base numbers when he asked. Checking Sensor Tower instead, he finds that "Gen AI referral numbers increased from 0.1% of traffic to 0.6% between Holiday 2024 and Holiday 2025" — roughly a 200% increase in absolute terms, smaller than Adobe's figure but still large on a percentage basis, while the share of total ecommerce traffic attributable to Gen AI referrals stayed minuscule.

Lipsman's method is to separate headline growth rates from the base they are computed against: "year-over-year growth rates like that tend to only occur from a very small base." He frames the episode as a case study in how an "anxious industry" amplifies unverified statistics — conference-season claims "amplified on LinkedIn" convert "speculation into conviction almost overnight," reinforced by consulting-firm forecasts citing EBIT risk and disintermediation. The Sensor Tower cross-check reframes the same underlying event — a real but small rise in AI-driven referrals — without the exaggerated growth-rate framing, and he credits Adobe's numbers as not "inaccurate" but lacking "important context."

This complicates Is non-human traffic really half of all internet use?, which measures a much larger agent-traffic share, but that figure counts AI crawler activity across the open web rather than consumer-facing referral traffic into retail sites specifically; Lipsman's point is that the metric chosen determines whether "agentic" traffic looks transformative or negligible. It also sits in tension with Will AI Overviews reduce search referral traffic to publishers?: publishers expect referral traffic to fall as AI answers replace clicks, while Lipsman's data shows Gen AI referral traffic rising, even from a near-zero base — the two findings describe different traffic flows (search engines diverting clicks vs. new AI-tool referrals into retail), and conflating them is part of the hype Lipsman is pushing back against.

The excerpt gives Lipsman's own characterization of the Sensor Tower numbers but not Sensor Tower's methodology or margin of error, so the comparison rests on his framing of a third-party dataset rather than on that dataset's primary documentation. It also does not establish that agentic commerce will stay small — only that the holiday-season growth-rate statistic, as reported, overstated the current scale of the shift relative to total ecommerce traffic.