If you read Part 1 of this series (What is AI), you now know that AI isn’t one thing. There are different types, different models, and different tools, and they’re not all built for the same jobs. Good. That knowledge is going to save you some headaches.
Now let’s talk about what to do with it.
Because here’s where it gets interesting. A lot of companies have already jumped into AI. They bought subscriptions, they set up accounts, some of them even put out a press release about it. And a surprising number of them are sitting here a year later wondering why it hasn’t moved the needle the way they expected.
The tool isn’t the problem. The lack of a plan is!
The “Just Get ChatGPT” Trap
I see this one a lot. A business leader hears about AI, does a little research, and decides the move is to get a team subscription to ChatGPT or a similar tool and hand it out to everyone. On the surface, that makes sense. It’s affordable, it’s easy to set up, and it feels like you’re doing something.
But here’s what usually happens. A few people on the team use it for writing emails or summarizing documents. A couple of people play around with it for a week or two. Then it quietly fades into the background because nobody had a clear reason to keep using it beyond “it’s there.”
That’s not an AI problem. That’s a strategy problem. There’s a big difference between having access to an AI tool and having a plan for how AI fits into your actual business workflows. The tool is just software. Without a strategy behind it, it’s just another app nobody uses.
I’ve watched this play out enough times now to know it’s not about the quality of the tool or the size of the company. It comes down to whether someone sat down and asked the right questions before buying anything.
What the Big Companies Are Actually Doing
It’s worth taking a look at how larger corporations are approaching AI, not because SMBs need to copy them exactly, but because it shows what’s actually working when companies get serious about it.
Amazon has built AI into nearly every layer of their operation. Their recommendation engine, their logistics network, their cloud services. They didn’t just hand employees a chatbot. They built custom systems trained on their own data, designed around their specific problems, and integrated them into workflows that already existed.
Google has done something similar. AI is baked into their search, their advertising platform, their productivity tools. They have dedicated teams whose entire job is figuring out where AI creates value and how to build it properly.
Salesforce has invested heavily in what they call Einstein, their AI layer that sits on top of their CRM. It does things like lead scoring, sales forecasting, and customer insights, all trained on the data their clients are already collecting.
The pattern here isn’t “big companies have better AI tools.” It’s that big companies have dedicated teams, clear strategies, and the resources to build solutions that are specific to their business. They’re not just using AI. They’re engineering it into their operations.
That’s a significant advantage. And if SMBs don’t start thinking along the same lines, the gap is only going to widen.
Where AI Actually Makes Sense for SMBs
You don’t need a team of data scientists or a seven-figure budget to get real value out of AI. But you do need to know which type of AI fits which problem. If you went back and read Part 1, you already have a head start here. Let’s put that knowledge to work.
Machine Learning: Smarter Predictions
Machine Learning is great for any situation where you’re trying to predict something based on existing data. For an SMB, that might look like forecasting which products are going to sell well next quarter, scoring leads so your sales team knows where to focus their time, or flagging customers who are likely to cancel so you can reach out before they do.
You probably already have most of the data you need sitting in your CRM or your sales platform. The question is whether you’re doing anything with it beyond basic reporting. Machine Learning turns that data into actionable predictions, and that’s where it gets valuable.
Generative AI: Content and Communication at Scale
Generative AI is the type most SMBs have already started experimenting with, and it’s genuinely useful when you know what to ask it to do. Writing first drafts of marketing copy, putting together proposal templates, summarizing long documents, building out internal knowledge bases so your team isn’t constantly asking the same questions. These are all solid use cases.
Where it falls short is when people treat it like a replacement for thinking. Generative AI is a drafting tool. It’s fast and it’s decent, but it still needs a human who understands the context to make sure the output actually makes sense for your business and your audience. I use it every day as part of my workflow. It speeds things up. It doesn’t replace the strategy behind the work.
Agentic AI: Automating the Stuff That Eats Your Time
This is where I think the biggest opportunity is for SMBs right now, and it’s also where most businesses haven’t even started looking yet. Agentic AI can handle multi-step tasks on its own. Customer service workflows that actually resolve issues without a human in the loop. Sales pipeline management that follows up, qualifies leads, and keeps things moving. Scheduling, onboarding, internal routing, the kind of repetitive work that takes up a huge chunk of someone’s day.
The key difference between Agentic AI and just using a chatbot is that an agent can take action. It doesn’t just answer a question. It does something with the answer. And when you build those agents around your specific processes, that’s when you start seeing real efficiency gains.
Strategy First, Tools Second
Here’s the order of operations that actually works. Define the problem first. Figure out which workflow is costing you time, money, or both. Then identify where in that workflow AI could step in. And only after you’ve done both of those things do you go looking for a tool.
Most businesses do it backwards. They find a tool they like and then try to figure out where to use it. That’s how you end up with a ChatGPT subscription that nobody opens after the first month.
Off-the-shelf tools are built for general use. They’re not designed around your processes, your data, or your specific problems. Sometimes that’s fine. But a lot of the time, especially as your needs get more specific, a custom-built agent or a fine-tuned model is going to do a better job. Custom doesn’t have to mean expensive or complicated. It just means built for your situation instead of everyone’s situation.
If you want a quick way to start thinking about this, ask yourself three questions. What’s a workflow in my business that takes more time than it should? Where in that workflow is a decision or a task being repeated over and over? And what data do I already have that could help AI do that task better? If you can answer all three, you’ve got the start of a real AI strategy.
Do You Need a Consultant or a Fractional AI Leader?
I want to be straight with you here, because I know how this section could come across. I run an AI consulting business. So yes, I have a stake in whether you think consulting is valuable. But I also talk to enough business leaders to know that most of them don’t need to hire a full AI team. What they need is someone who understands both the technology and their business well enough to bridge the gap.
That’s what a fractional AI role or a consulting engagement actually looks like in practice. It’s not someone coming in to sell you software. It’s someone who can look at your specific operation, figure out where AI actually adds value, and help you build it in a way that makes sense. Someone who’s going to tell you when the answer is a simple tool, and when it’s something more custom.
The cost of guessing your way through this is higher than most people realize. Not because AI is dangerous or complicated, but because a bad implementation wastes time, wastes money, and makes the whole team skeptical of the next attempt. Getting it right the first time, or at least close to right, is worth the investment.
If you’ve got questions about where AI fits in your business, I’m happy to talk it through. No obligation, no pitch deck, just a conversation about what makes sense for your situation.
The Bottom Line
AI is only as useful as the strategy behind it. The tool matters a lot less than the plan. The companies that are getting real value out of AI right now aren’t the ones with the flashiest tools. They’re the ones that took the time to figure out where AI actually solves a problem for them, and then built around that.
You don’t need to become an AI expert overnight. You just need to stop treating it like a one-size-fits-all solution and start asking better questions about where it fits in your business. That’s where it starts.
Great Lakes AI Solutions helps small and midsize businesses across the Midwest build AI strategies that actually work for their operations. If you want to talk about what that looks like for your company, get in touch. And if you haven’t read Part 1 yet, start there.
