Toast IQ Logged 21 Million Conversations Last Quarter. Sales Questions Outnumbered Labor Questions Nearly 4 to 1.

Toast IQ, the AI assistant built directly into the Toast POS, logged 40,000 weekly active users and more than 21 million conversation threads in Q1 2026. Sales and revenue questions were the single largest query category at 47% of restaurants, ahead of labor-cost questions at 13% and voids/comps/loss-prevention questions at 12% — and fine-dining locations averaged 195 threads each, about 29% more than fast-casual operators. The assistant isn’t limited to answering questions: operators can adjust pricing or mark items out of stock directly from the chat interface, collapsing what used to be a settings-menu task into a conversation.

The volume matters more than the novelty here. Toast built its own Toast IQ Grow feature — AI-driven marketing paired with a dedicated human marketer — around this same engine, and cites Matoi’s 4.5x return on ad spend and Flava Wings’ 3.5% year-over-year January sales increase as results. Toast is positioning the product directly against Square’s Managerbot, with the stated differentiator being that Toast IQ runs natively across the full platform rather than as a bolt-on.

Toast’s own earlier robotics roundup is a useful check on how the company thinks about sequencing this kind of investment. That piece — a survey of ten real-world restaurant robot deployments, from Chipotle’s Autocado prep robot to Sweetgreen’s 500-bowl-per-hour Infinite Kitchen makeline — explicitly frames Toast IQ as the decision layer that should come first: use data to identify where prep, throughput, or front-of-house work is actually bottlenecked, then evaluate whether a specific piece of hardware solves that problem. Toast’s argument, in effect, is that the 21 million conversation threads aren’t just a usage number — they’re the diagnostic layer that tells an operator whether the next dollar belongs in a labor-saving robot or somewhere else entirely.

For an operator evaluating AI investment, the split matters: a platform where sales questions dominate over labor questions is being used as a revenue tool first and a cost-control tool second — worth checking against how a given property’s own usage would likely break down before assuming the reverse.