Michaels Launches 'Ask Mike' AI Shopping Assistant on Gemini

What happened

Michaels, the US arts-and-crafts retailer, has launched "Ask Mike", a generative AI shopping assistant built on Google Cloud's Gemini Enterprise platform. The tool is designed to help shoppers move from an open-ended creative idea to a specific, confident purchase by interpreting vague project intent and translating it into product recommendations.

According to Google Cloud, the assistant sits within Michaels' customer experience strategy as a way of supporting shoppers who arrive with inspiration rather than a defined shopping list — a common pattern in the crafts category — and guiding them through the decision-making process conversationally.

Why it matters

Crafts retail sits at an unusual intersection of high inspiration and high decision friction: shoppers often know what they want to make but not which materials, quantities or techniques it requires. That gap between intent and action is a classic behavioral-economics problem — choice overload and uncertainty routinely cause cart abandonment or under-buying. A conversational assistant that narrows options based on described intent directly targets that friction point rather than simply speeding up search.

For service design teams, this is a useful signal that generative AI is being positioned less as a chatbot for support queries and more as a guided decision-making layer embedded in the discovery journey itself — reshaping how retailers think about the earliest, fuzziest stage of the customer journey.

The Renascence take

The interesting part of "Ask Mike" isn't the technology — it's the job it's been given. Most retail AI assistants are deployed to answer questions faster; this one is deployed to help shoppers figure out what they actually want, which is a fundamentally different and harder design problem.

Most brands bolt generative AI onto customer service and call it innovation. Michaels is using it earlier in the journey, at the point where intent is still fuzzy and shoppers are most likely to freeze or walk away — that's the higher-value moment to intervene. The lesson for operators isn't "add a chatbot," it's "find the specific decision point where your customers get stuck, and design the AI experience around resolving that friction, not around deflecting contact volume."