A friend who runs a small landscaping company out near Wayzata told me last month that he'd been pitched seven different AI tools in the previous two weeks. Some came through email. Some through LinkedIn. Two were demoed in person by sales reps who had clearly memorized a script. He asked me which one was actually worth paying for.
The honest answer was that most of them were the same product with different names. And none of them addressed what was actually broken in his business.
That conversation isn't unusual. Owner-operated businesses around the Twin Cities are getting hit with AI pitches at a rate that didn't exist eighteen months ago. Some of those tools do something useful. A lot of them are dressed up versions of things you already pay for, with "AI" stuck on the front end.
If you're trying to figure out what's worth your time, the question to start with isn't which tool. It's which problem.
What most of the AI pitches look like
Walk through an exhibit hall at a small business expo right now and you'll see roughly the same product, repackaged, at twenty different booths. There's an AI chatbot for your website. There's an AI assistant that writes your social media. There's an AI tool that summarizes your meetings. There's an AI dashboard that pulls metrics from your other tools.
A lot of these are real products built on real technology. A few of them might even be useful for your specific business. But here's what they have in common: they're sold as a category, not a fix. The pitch is "here's a tool, see if you can find a use for it." Not "here's a problem, here's how this addresses it."
That's backwards from how a small business should buy software. You don't have time to figure out where a tool fits. You have a problem. The tool either solves it or it doesn't.
The tools that get a lot of attention but rarely deliver
I'll be specific about a few categories where I see business owners spending money for limited return.
AI website chatbots. These get sold hard because they're easy to demo. The problem is that most of them don't know anything specific about your business. They answer the questions a generic chatbot can answer. The actual high-value questions, the ones a customer would normally call about, get handed off anyway. If you're going to install one, the real work is training it on your services, your pricing, your scheduling. That work usually doesn't happen, so the chatbot stays generic and the customer learns to ignore it.
AI content generators for social media. These produce output that looks fine and feels off. Customers can tell. The point of social media for a small business is to sound like the actual person behind the business. A generated post is the opposite of that.
AI dashboards and analytics layers. These take data you already have and put a chat interface on top of it. They don't change what you do. If you weren't reviewing the numbers before, you won't review them now just because the report has a chat box.
None of these are scams. They're real tools. They just don't usually move the needle for an owner-operated small business.
The tools that actually help
The AI work that produces results for small businesses is almost always invisible. It runs in the background. The customer doesn't know it's there. The owner notices that something that used to fall through the cracks no longer does.
A few examples from work I've actually done.
A lead inquires through a contact form. They get a personal-sounding response within five minutes, with a link to book a call. If they don't book, they get a thoughtful follow-up two days later. Then another a week after that. The owner doesn't write any of those messages in the moment. They were written once and now run on their own.
A patient finishes a course of physical therapy. Six months later, they get a check-in note asking how things are holding up. If they reply that something's flaring back up, the practice gets a nudge to call them. If they don't reply, nothing happens. Some percentage of those messages turn into reactivated patients.
A home inspector finishes a report. The agent who referred the inspection gets a brief, useful note three months later. Not a marketing blast. Not a newsletter. A short, specific message that keeps the relationship warm. Most inspectors I talk to say they meant to do this for years and never got to it.
None of these involve a flashy AI product. They're a handful of automated messages, written carefully, triggered at the right moment. The AI part is mostly about generating personalized text and routing things based on what comes back. It's the smallest possible piece that fixes the actual problem.
How to tell the difference
When someone pitches you an AI tool, the test isn't whether it works. It's whether it addresses something that's currently broken in your business.
Ask yourself two questions. First, what's currently falling through the cracks. Leads going cold, customers drifting away, documents not coming back, referral sources going quiet. Second, would this specific tool fix that specific thing.
If you can't draw a clear line from the tool to a current problem, you don't need the tool. You're being sold a category. Categories don't fix anything.
The boring truth is that most small businesses don't need anything new. They need to turn on the part of what they already have that handles follow-up, or build a small system that does it. That's not as exciting as a new AI product. It works better.
If you want to talk through what this would look like for your business, the audit is free and takes 30 minutes.