From keyword search to conversational AI shopping
Pinterest’s Ask Pinterest app is an experimental conversational AI shopping tool that lets people discover products and ideas through natural dialogue instead of traditional keyword-based search, using context, taste signals, and multi‑step chat to guide them from inspiration to purchase decisions. This limited-access web app runs separately from the main Pinterest experience, giving the company room to test new interaction patterns without overhauling how existing users browse today. Rather than typing short, transactional queries, people can describe broader goals, such as planning a dinner party or furnishing a room over time, and receive tailored suggestions in response. Pinterest frames this shift as a new “query shape”: discovery driven by ongoing conversation rather than one-off searches. By blending conversational AI shopping with Pinterest’s visual discovery heritage, the company aims to support complex journeys that unfold over days or weeks, not a single session.
How the Ask Pinterest app personalises AI product discovery
Ask Pinterest is built on Pinterest’s Taste Graph, an internal map that connects users to interests, aesthetics, and styles drawn from their activity on the platform. Instead of generic results from across the web, the app grounds answers in Pinterest-native signals such as saved Pins and Boards, which it can use when a user is signed in. This allows the conversational AI to understand that someone who saves minimal, neutral-toned interiors will likely want different furniture ideas from someone collecting colourful maximalist designs. The system is also designed to retain context across sessions, so it can keep track of evolving plans, like refining a wedding mood board or upgrading a home office. In practice, Ask Pinterest turns AI product discovery into a personalised, iterative dialogue that reflects declared taste, observed behaviour, and ongoing intent rather than a fresh start each time a user opens the app.
Business Assistant, MCP and Performance+: new Pinterest shopping tools
Alongside Ask Pinterest, the company is rolling out new Pinterest shopping tools for advertisers that use AI to automate planning and optimisation. Business Assistant, currently in closed beta in Ads Manager and mobile, acts as an AI collaborator that highlights trends, performance status, and content opportunities using visual outputs like graphs and recommended Pins. Pinterest’s Model Context Protocol (MCP) is an infrastructure layer that connects the platform’s campaign, analytics, and keyword insights to external copilots and agentic tools. According to Pinterest, MCP is being developed with partners including PMG, Pacvue, Dentsu, Havas, Innovid by Mediaocean, and Omnicom’s Jump450. Performance+ creative adds a new AI model that chooses among multiple creative variations and, in testing, increased click volume by 7.5% compared with a single-variant approach. Together, these updates aim to reduce manual campaign work while improving how ads match user intent.
What AI-powered retail discovery means for shoppers and brands
Pinterest’s recent launches point to a strategic pivot towards AI-powered retail discovery where context and taste matter more than keywords. Chief business officer Lee Brown states that “the future of discovery won’t be driven by keywords alone,” positioning Pinterest as a platform where people come to plan, curate, and act on what they want to do next. For shoppers, Ask Pinterest offers a more guided, conversational way to move from inspiration to action, especially in longer decision journeys like home projects or event planning. For retailers and advertisers, Business Assistant, MCP, and Performance+ shift workflows toward AI-guided optimisation, with campaign data flowing into third-party tools and creative selection becoming more automated. As conversational AI shopping becomes more common, brands will need assets and content that support multi-step dialogue, not just single-click conversions, to stay visible inside these emerging chat-based discovery journeys.
