Why Hakuhodo DY’s “AI Search Usage Survey” hits a nerve for online retailers

Reviewed: ad hoc news Accessory & Components desk. Edited and checked on 2026-06-17, 22:04. Details in the imprint.

Hakuhodo DY ONE’s “AI Search Usage Survey” is one of those quiet products that does not sit on a shelf, but on the desks of marketers who suddenly feel very seen. It turns fuzzy talk about AI hype into crisp percentages that hurt a little - in a good way.

What the survey actually measures

Hakuhodo DY ONE positions the AI Search Usage Survey as a practical tool for brands that suspect their customers are already using ChatGPT, Gemini or similar tools during the purchase journey, but lack hard data. The latest wave focuses on how often and for what people use AI search around shopping.

According to reporting on the study in Japanese specialist outlet Markezine, more than 70 percent of users say their frequency of AI search use has increased, with a clear tilt toward product research and comparison before checkout. For retailers this means AI is no longer an edge case, but a parallel search layer they cannot see.

How Hakuhodo DY ONE packages the insights

The AI Search Usage Survey is not a glossy PDF for the public. It is a recurring, data-heavy research product that Hakuhodo DY ONE’s consultants use in workshops and pitches with e-commerce clients, often combined with site audits and UX reviews. You can imagine the slide: heatmaps, funnel charts, blunt bullets.

In practice, agencies use the numbers to argue for concrete measures on clients’ sites. That ranges from clearer product copy and FAQ structures that play nicer with AI summarisation to content that answers the sort of exploratory questions people type into AI tools rather than into Google.

Where the findings hit online shops

The sobering part for shop operators is simple: a user may read your lovingly crafted landing page only in distilled form, via an AI answer box. Hakuhodo DY ONE’s survey underlines that many respondents treat AI as a “first filter” before they even click through to a brand site.

That changes what “good” content means. It is no longer enough that a human reader finds a paragraph clear. The copy also needs to be structurally robust, with unambiguous product specs, clear pricing logic and explicit answers to typical objections, so AI systems can extract and relay them reliably.

Why a media group cares this much

For Hakuhodo DY, the survey is more than a marketing gimmick. The group has long positioned itself as a partner for integrated communication, where media planning, creative work and data strategy are tightly woven. Understanding AI search behavior is now a structural part of that package.

Subsidiaries such as Hakuhodo DY ONE sit exactly at the intersection of performance marketing, CRM and site optimisation. Regularly updated proprietary surveys give them a differentiator when they pitch against global networks and consulting houses that rely heavily on third-party research.

What surprises in the user behavior

One detail from the coverage of the AI Search Usage Survey will prick up many ears in e-commerce: a notable share of users say they trust AI summaries more when comparing similar products than they trust individual brand pages. That trust is not necessarily rational, but it is a fact to work with.

Another point is the role of AI in “almost abandoned” sessions. Some respondents report copying long cart contents or spec lists into AI tools to double-check compatibility or value just before paying. That quiet last-minute consultation can tip the decision toward or away from a retailer.

Practical takeaways for retailers

What do you do with such a survey if you run a midsize online shop? In Hakuhodo DY ONE’s typical consulting projects, the answer starts with content hygiene: unifying spec tables, cleaning up duplicate product descriptions, and structuring Q&A blocks so they read like self-contained mini-answers.

The agency also pushes for better tracking of off-site behavior, for example by watching referral patterns and query strings in site search to infer when AI tools are likely in the loop. Combined with the AI Search Usage Survey, this helps clients prioritise which categories and funnels to fix first.

How it compares to other research products

Globally, many consultancies publish thought-leadership on AI adoption. What makes Hakuhodo DY ONE’s AI Search Usage Survey stand out is its narrow focus on concrete usage around commerce and the Japanese digital market, where mobile and super-app habits differ from Western benchmarks. That specificity is a strength.

At the same time, the core patterns - more AI use, more trust in AI summaries, AI as a pre-search filter - resonate with findings from international research houses that follow generative AI uptake across markets. For multinational retailers, the survey thus offers a grounded local lens that still maps onto broader trends.

Group context and stock angle

For Hakuhodo DY Holdings, products like the AI Search Usage Survey play into a larger strategy of layering data and consulting services onto its traditional agency business, making it less dependent on pure media volumes. The quiet product in the slide deck is part of a louder shift in the business model.

Shares of Hakuhodo DY Holdings Inc (JP3768600003) trade on the Tokyo Stock Exchange in Japanese yen.

This article was AI-assisted and editorially reviewed. Product information without guarantee; prices and availability may change at short notice. No investment advice, no buy or sell recommendation. Stock-market transactions involve risks up to total loss.