Ask Mike: Michaels Launches Gemini AI Shopping Assistant for CX

What happened

Michaels, the North American arts-and-crafts retailer, has launched Ask Mike, a generative AI shopping assistant built on Google Cloud's Gemini model, designed to help customers move from a vague creative idea to a specific product or project plan. The tool is live across Michaels' digital channels and represents one of the more substantive deployments of enterprise-grade generative AI in the speciality retail sector.

Ask Mike is intended to handle the distinctly open-ended nature of craft shopping — where a customer might arrive knowing only that they want to "make something for a birthday" — and guide them through product discovery, material selection and project inspiration in a conversational format. The assistant draws on Michaels' product catalogue and project content to surface relevant recommendations rather than returning generic search results.

Why it matters

Craft retail sits at an unusually complex intersection of intent and discovery. Unlike commodity shopping, where a customer knows the SKU they want, craft purchases are frequently driven by aspiration rather than specification. That ambiguity has historically been resolved by knowledgeable in-store staff — a resource that does not scale online. Ask Mike is a direct attempt to replicate that consultative, needs-clarifying conversation at digital scale, which is precisely the service-design problem that large language models are structurally well-suited to address.

From a behavioural economics standpoint, the assistant also works to reduce choice overload — a well-documented friction point in categories with thousands of SKUs and no obvious decision hierarchy. By anchoring the interaction around the customer's creative goal rather than a product grid, Michaels is effectively shifting the cognitive load from the shopper to the system, which research consistently links to higher conversion and satisfaction. For CX practitioners, this is a live case study in using AI not merely to automate queries but to restructure the discovery journey itself.

The Renascence take

Most retail AI announcements focus on speed and deflection — answering FAQs faster, reducing contact-centre volume. Ask Mike is framed differently, as a creative collaborator rather than a cost-reduction tool, and that framing distinction matters more than the underlying technology.

The risk most operators will miss is that a generative assistant trained primarily on catalogue data can easily become a sophisticated search bar in disguise — fast, but not genuinely consultative. The behavioral principle at stake is goal-based framing: customers who articulate a desired outcome before browsing make more confident, higher-value purchases and report greater satisfaction. Michaels' real design challenge is ensuring Ask Mike consistently elicits that outcome statement rather than defaulting to keyword matching. Operators considering similar deployments should instrument for goal-capture rate, not just query resolution rate — because resolving the wrong question quickly is still a failed experience.