Business 4 min read By Callum Montgomery
Big Tech racing for small business AI assistants, but customers still come from referrals
Meta, Google, OpenAI and Anthropic are competing to become the AI assistant of small business owners, but the author argues the tools will save time rather than win new customers, which still arrive mostly through personal referrals.
Meta, Google, OpenAI and Anthropic are competing to become the artificial intelligence assistant of small business owners, and the tools they are building promise to take over administrative work. Meta launched its Muse agent for small businesses in September, while Anthropic and OpenAI are packaging owner-focused tools and Google is embedding features into software that businesses already use. The author, who runs Alignable, a network of about 10 million small business owners, argues that these products will save owners time but will do little to help them find customers.
The central point of the argument is that customer acquisition for small firms depends on relationships rather than software. Alignable members report that about 85 per cent of their customers arrive through referrals. The network has formed 226 million business connections across 35,000 communities. The author summarises his own experience in a rule of thumb: 50 warm relationships are worth more than 500 cold contacts. He is clear that this is an experience-based heuristic rather than a measured result. The company's data suggests a typical member has about 190 ideal customers who are one introduction away, and one in four has more than 600.
The most valuable referral partners, according to the author, are complementary businesses that serve the same customer rather than direct competitors. A florist, a wedding photographer and a caterer, for instance, can each send work to the others without competing for it. Software that generates lists or sends messages does not, on its own, create these trust-based links between owners who already share customers.
Evidence on the productivity gains from AI tools is mixed. A 2026 study by Bluehost of 350 US owners found that 78 per cent said AI saves them time each week. The same study found that 43 per cent reported no change in revenue. The author cites an example from Paul Birrell, whose automation sent about 400,000 emails in three days. Volume of outreach, in other words, does not automatically translate into sales.
The author proposes dividing AI tools into three types: a personal assistant that manages an owner's day, business-system AI that connects to financial and sales software, and relationship-network AI that helps owners identify and deepen referral connections. Meta's Muse, for example, connects to QuickBooks, Shopify and Stripe, which places it in the business-system category. The author is open about a conflict of interest, noting that Alignable co-hosts small business AI training with Anthropic.
The scale of the sector makes the question significant. There are about 36 million small businesses in the United States, and more than 80 per cent have no employees. For these owners, time spent on bookkeeping or email is time not spent building relationships. The author raises a concern that companies funded by advertising may steer owners towards buying advertising, and he questions whether AI assistants will ultimately serve the owners who use them or the ad-driven platforms behind them. That question, he suggests, remains open.
Source: fortune.com
3
