Gusto opens its multi-client MCP to every accounting firm while the rebuilt Gusto Pro stays in early access

Before this release, a firm that wanted an AI tool working on its clients' payroll made one connection per client and reconnected each time it switched. Now it makes one connection for the whole book ^[11]. Bridgette Haymaker, Gusto's general manager and vice president of product management, described the old setup this way: "Until now, working with AI across a client book has meant connecting to one client at a time and disconnecting and reconnecting to switch." ^[3]

We think the standing routine is the useful part. Haymaker's own example is a task that runs every Thursday, reviews client emails for payroll hours, preps each client's payroll and notifies the accountant when it is ready for review ^[5]. The email reading and the keying move to the agent, and the approval stays with the accountant. Haymaker wrote: "Every change still comes back to the accountant to review and approve before anything is updated in Gusto, and only the clients they've chosen are ever connected." ^[4]

That puts a ceiling on the saving, and the ceiling is review time. A firm recovers its prep hours minus whatever it spends re-checking what the agent pulled from emails. A firm that checks every extracted hour against the original message gets back little. A firm that spot-checks gets back more but takes on the extraction error rate. The report on the announcement does not include an error rate, a price for the MCP or any adoption figure.

Gusto is building for two surfaces. In an Oct. 7 blog post Haymaker said firms can put agents to work on payroll prep "wherever they already are, whether that's in Gusto Pro or in the AI tools they use every day" ^[13]. Gusto is not asking firms to move their AI work into its own window. The Partner Program beside it is free and tiered, and it pays firms a revenue share as they grow their practice on Gusto ^[9]. The price to the firm is zero, and the value to Gusto is presumably more client books running payroll on its platform (our inference, not a Gusto statement).

We could be wrong in two ways. If firms in early access report that review takes as long as prep used to, the saving is gone. If the hours do come back but nobody sells advisory work with them, the product has saved time without moving the firm up the value chain, and Gusto says the updates are designed to help firms move from transactional tasks toward higher-value advisory services ^[12].