Michaels says its Google-powered AI assistant doubles the conversion rate of traditional search

Michaels' AI assistant Ask Mike converts shoppers at double the rate of traditional search. The tool drives a 27% add-to-cart rate across 75,000 conversations.

Categorized in: AI News Marketing
Published on: Jul 27, 2026
Michaels says its Google-powered AI assistant doubles the conversion rate of traditional search

Michaels lifted the curtain on July 21 on its new AI shopping assistant, Ask Mike, revealing that the tool has quietly logged 75,000 conversations since May and is converting online shoppers at more than double the rate of traditional keyword search. For marketing teams watching the retail AI landscape, the numbers offer a concrete benchmark: conversational product discovery can directly lift purchases and average order value.

Ask Mike lives on the crafts retailer's desktop and mobile website, as well as its iOS and Android apps. Customers type natural-language prompts such as "Help me plan my kid's birthday party" or "I need fabric for DIY curtains," and the assistant asks follow-up questions about colors, materials, project type, and budget before serving a curated set of products. It was built using Google Cloud's Gemini Enterprise for Customer Experience, the same software package used by Ulta Beauty, Macy's, and The Home Depot for AI-powered shopping functions.

Doubling conversion through guided discovery

Shoppers who interact with Ask Mike convert - making a purchase online after the conversation - at a rate double that of customers who rely on the site's standard search bar, according to Heather Bennett, president and chief customer officer at Michaels. The company also reports that 27% of Ask Mike sessions result in a click on a product link or adding an item to cart.

"Unlike traditional search, Ask Mike acts as a creative assistant that helps shoppers move seamlessly from inspiration to product discovery and cart action, driving much higher overall engagement," Bennett said. "Traditional search requires customers to break projects down into isolated keywords like 'floral fabric', leaving them to sift through the results themselves."

Bundling, not just finding a single item

Bennett observed that users of the AI assistant frequently bundle core supplies and finishing touches within a single session - for example, everything needed for a themed party, a framing project, or a craft. Traditional search users, she said, more often hunt for one specific item they already had in mind. That behavioral difference suggests Ask Mike is expanding the moment from "I need glue" to "help me complete a project," increasing the number of items per transaction.

From concept to production in six weeks

Google Cloud's global director of retail industry strategy, Paul Tepfenhart, noted that Michaels moved from concept to a production-grade AI tool in just six weeks. Kapil Dabi, Americas lead for retail and CPG solutions at Google Cloud, attributed the speed to the retailer's existing cloud infrastructure on Google Cloud and to a product catalog that was already organized for AI retrieval.

"Beyond the technical foundation, I would say success also requires a clear point of view on the customer experience," Dabi said - knowing, for instance, that customers planning a birthday party will need specific types of products at specific moments. He added that retail is entering an era where conversational AI becomes the primary interface between shoppers and brands, with discovery that is faster, more intuitive, and personalized.

Why this matters for marketing professionals

The Ask Mike results illustrate that moving from keyword search to a conversational, consultative assistant can be a measurable lever for AI for Marketing strategies and a direct driver of AI for Sales outcomes. For marketers evaluating similar tools, the metrics worth tracking go beyond raw conversion rates: add-to-cart clicks (27% of interactions in Michaels' case) and session-level bundling become early indicators of whether an AI assistant is expanding basket size, not just streamlining search. The six-week development timeline also signals that building on a clean, API-ready product catalog can accelerate time-to-value for generative AI commerce projects.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)