Executive summary: On November 3, 2026, Meta expanded its Muse AI agent to small business, adding a set of connectors that plug Muse into the tools owners already run — Shopify, QuickBooks, Stripe, Canva, Slack, Zoom, Klaviyo, Dropbox, Notion, Figma, Asana and Box — along with their Instagram professional accounts, Facebook Pages and Meta ad accounts. The expansion sits on a base Meta itself sized on its October 21, 2026 earnings call: finance chief Susan Li said Meta estimates at least 500 million self-identified small businesses use its family of apps. The move follows the day earlier, when Meta launched the Meta Enterprise Platform and hired Chirantan "CJ" Desai, until then the sitting CEO of MongoDB, to lead it, reporting directly to Mark Zuckerberg. The headline for owners is not "another chatbot." It is that the agent now has a connect-your-tools layer and a human-approval gate: publishing, sending, or spending still requires your sign-off. This playbook gives four steps to wire one connected agent into the one job that makes you money, with a human gate on the money step, this month — all of it free to start.
The obstacle here is not intelligence; it is back-office hours. Small businesses are short on hours, not ideas — they have the ads to write, the direct messages to organize, the inventory to keep current, the invoices to chase, but no AI team to own it. Most owners in the 500-million base are already typing into a general chatbot every day, but that window has no memory of your price sheet, your store, or your customers. So the first move is scope, not spend: name the single workflow that makes you money and that you currently lose time on. For a product business it is usually the order confirmation, the reorder prompt, or the ad that brings in revenue. For a service business it is the estimate or the follow-up. One job. If you spread across five connectors on day one, you will measure none of them and integrate none of them.
Tools & pricing: A notebook. The decision costs nothing; the mistake is picking the fun job (brand graphics) over the money job (the ad or the invoice that converts).
The turn in this story is the connect-your-tools layer. Muse launched for small business in late September as "a set of new skills and connectors inside the existing Muse app," and the November 3 expansion widens that list to the tools owners already run. That is the missing piece most agent products do not have: an agent that can read your store, your books, and your brand instead of starting from a blank prompt. So take your one money job and connect the single system where that job lives. If your money job is orders, connect Shopify. If it is getting paid, connect QuickBooks and Stripe. If it is marketing, connect your ad account and Canva. Give Muse the goal — "run my ads," "keep my site current with new inventory," "organize my direct messages" — and let it draft. Resist the urge to wire up the whole company on day one; the value of a connected agent is depth on one job, not breadth across twelve.
Tools & pricing: Muse for Small Business, free for most functionality with usage limits; a subscription is available for higher tiers. Connectors named by Meta: Shopify, Stripe, QuickBooks (Intuit), Canva, Slack, Zoom, Klaviyo, Dropbox, Notion, Figma, Asana, Box, plus Instagram professional account analytics, Facebook Pages and Meta ad accounts. Verify current availability and eligibility directly with Meta.
This is the "why now" and the credibility in one. Meta has been explicit that actions such as publishing, sending, or spending still require the user's approval. The agent drafts; you approve. For an owner who was burned by the idea that "an agent that acts is an agent you cannot control," that rule is the whole point. The owners who win this round are not the ones with the fanciest model — they are the ones who deliberately configure the gate: which drafts get a look, which sends need a click, which spending is out of bounds. Set that rule before you let the tool run, not after you have to catch it. This is the same discipline the in-suite agents have exposed all along; the difference is that Muse now pairs it with the connectors from Step 2, so the gate sits on a real workflow instead of a blank chat.
Tools & pricing: Your own two rules: an approval threshold for anything outbound, and a hard stop on anything that moves money. The discipline, not the dashboard, is the deliverable.
Three quarters of the owners in Meta's base are small or solo, and the realistic outcome of a connected agent is not "it runs my business for you." It is the hours you were paying an outside service to reclaim. So measure the one job for two weeks: how long it took by hand, how long it takes with Muse drafting, and how many of those artifacts actually convert to revenue. That one number — not the tool's logo — decides whether you scale it, add the next connector, or pull back. The broader context that makes this worth testing now: Meta called the enterprise push "the next major pillar" of its business, hired a database CEO to run it, and said it envisions consumer agents and business agents eventually connecting and transacting directly. Owners who have a measured, gated, connected workflow today are the ones positioned when that layer arrives.
Tools & pricing: A two-week measurement. Track three fields — baseline time, assisted time, conversion — for the single job.
The shift is not "adopt AI" — half a billion small businesses are already in the room. It is "connect it." The gap between a chatbot and a working agent was never a smarter model; it was the connect-your-tools layer plus a human gate on the money step. Point one connected agent at the one job that makes money, keep a human on the money step, and measure before you scale. The realistic ceiling is not automation-for-its-own-sake. It is the estimate, the ad, and the invoice moving from your desk to a tool that already knows your store — at a near-zero cost to start — so the jobs your competitors are still doing by hand are the ones you close first. That is how the owners who connect and gate end up ahead of the ones still running AI out of a chat window.
Your next twenty minutes should be a scope decision, not a strategy debate. Write down the one job that makes you the most money. Pick the single tool where that job lives and connect it. Start the free "Grow with Google" AI certificate this week if you want a structured refresher. Set the "human on the money" rule. Then measure for two weeks. To get the full SMB agentic-AI strategy — including how a connected agent like Muse stacks with in-suite agents and where the free tier actually pays for itself — start here:
Get the SMB AI adoption strategy