On 15 June, Salesforce agreed to buy the AI customer-service firm Fin for $3.6 billion. The message for small business is blunt: an AI agent that answers your customers around the clock is no longer experimental.
Something quietly important happened in the world of business software this week, and it tells you a lot about where customer service is heading. On 15 June 2026, Salesforce, the company behind the customer software a huge share of the world's businesses run on, agreed to buy an AI customer-service firm called Fin (formerly Intercom) for $3.6 billion. TechCrunch reported the deal, which Salesforce says it will fold into Agentforce, its platform for building AI agents that handle real work.
On its own, a corporate acquisition is not blog-worthy news for a cafe or a trades business. What matters is the signal underneath it. When the largest customer-software company on earth pays more than three billion dollars for an AI that answers customer queries, it is not betting on a gimmick. It is confirming that a category most owners still think of as flaky chatbots has quietly become genuinely good, and worth a great deal of money.
For a small business, the price tag is not the interesting part. What is interesting is what these agents can now actually do, and how close that capability has come to being within reach of an ordinary business rather than just the big end of town.
Forget the old idea of a chatbot that offers three buttons and a dead end. Fin, the firm Salesforce just bought, runs an AI agent that resolves customer enquiries across the channels people actually use: live chat on a website, WhatsApp, SMS, even phone calls. The difference from the chatbots of a few years ago is that it understands a real question asked in plain language, answers it properly, and only passes the conversation to a human when it genuinely needs to. It is less an automated menu and more a capable first responder who never sleeps.
That is the leap. The reason these tools failed for so long was that customers could tell within one message they were talking to a machine that did not understand them, and it was infuriating. The current generation crosses a quality line the old ones never did, which is exactly why a buyer like Salesforce is willing to pay billions to own one rather than wait for the technology to mature.
Deals like this are a useful tell for a small business owner trying to work out what is hype and what is real. The companies with the most to lose from getting it wrong are voting with their cheque books, and they are voting for customer-service agents. Ethan Mollick, the Wharton professor whose writing on AI at work many owners already follow, has made the point repeatedly that the businesses pulling ahead are not the ones with the fanciest tools, but the ones who rethink which tasks they hand to AI in the first place. Mollick is talking about exactly this kind of work: a whole job, answering customers, handed over rather than a single clever reply.
The good news is that this capability does not stay locked inside billion-dollar acquisitions. The same underlying technology is what powers the customer agents a small business can put on its own website and channels today. The gap is no longer whether the tools are good enough. It is whether they are set up to represent your particular business well, which is a different problem entirely.
Think about where you lose work right now. It is rarely a customer who shopped around and chose someone else. Far more often it is the enquiry that came in at 9pm and got answered at 9am the next day, by which time they had already booked the competitor who replied first. It is the message that arrived while you were elbow-deep in a job and slipped down the list. It is the same five questions about hours, pricing and availability that eat your evenings. Every one of those is revenue quietly leaking out the back, and it is exactly the leak a good customer agent plugs.
Done well, an AI agent does not replace the human warmth that makes a small business worth dealing with. It protects it, by catching everything that would otherwise fall through the cracks and freeing your people to spend their attention on the customers in front of them. Here is what good actually looks like once one is set up properly:
The question is no longer whether an AI can answer your customers well. The big buyers have settled that. The question is whether yours is set up to sound like you and protect the revenue you are quietly losing.
None of this works by flicking a switch. An agent that answers customers badly is worse than no agent at all, because it puts your brand's voice in front of people when it is wrong. The craft is in the setup: feeding it the right knowledge about your business, tuning its tone so it sounds like you and not a call centre, deciding precisely when it should step back and let a human take over, and wiring it into the tools where your bookings and orders actually live. That is the same lesson we wrote about in our guide to AI agents that actually help, and it is the difference between an agent that wins you work and one that quietly costs you customers.
The technology has crossed the line from experiment to proven, and the biggest players in software have just confirmed it with their wallets. The advantage now goes to the businesses that put it to work first, and put it to work properly. That is precisely what we do at NextAura: we build and run the customer agents that answer your enquiries around the clock, in your voice, and hand the right moments to your team. Get in touch and we will set it up properly while you get back to looking after the customers in front of you.
Tell us where you are headed. We will come back with a scope, a price, and a launch date you can plan around.
Book a free consultation
AI AgentsGoogle's new Gemma 4 12B is small enough to run on an ordinary laptop, and a wave of developers say local models are finally good enough for real work. For a small business, that means powerful AI without the data leaving the building.
AI AgentsGoogle has connected your Business Profile to its Gemini app, so AI can draft review replies, update your hours, summarise your performance, and flag gaps. Here is what it means for your local business.
AI AgentsThe gap between an impressive demo and a dependable agent is where most projects stall. Here is how we design agents that survive contact with real customers.