For almost two years now, my team at Impact and I have been in conversations with small and mid-sized businesses about AI. Workshops. Discovery calls. Strategy sessions. Conferences. Casual conversations that turned into something deeper. Hundreds of them.
I went in with assumptions. Some held up. Most didn't.
What I want to share here isn't a pitch. We're not ready for that yet, though it's coming. What I want to share is what we've heard. Because if you're a business leader trying to figure out where AI fits, you're probably not alone in what you're feeling.
Hear more about what our experts have found out in the AI field by watching Impact’s webinar, Lessons Learned from Incorporating AI into Business Processes.
The Efficiency Trap
When people talk about what they want from AI, the word that comes up most is efficiency. Less heavy lifting. Faster processes. A world where things just work.
I get it. But here's what I've learned: most processes aren't worth optimizing. They're full of workarounds, inherited dysfunction, duct tape, and habit. Making them 20% faster doesn't change the output. It just gets you to the same place a little quicker.
Just recently, I spoke with an AI engineer who'd been handed a project to speed up part number lookups in sales orders. The idea was straightforward: make the process faster. But as we talked it through, the real question became obvious: why optimize a process that shouldn't exist in that form at all?
What if the AI moved upstream, into the quoting process itself, making the sales order aware of inventory, pricing structures, and lead times from the start? What if customers could get real-time updates on configurations based on timing and cost? That's not efficiency. That's a different way of working.
I've seen the same shift in HR teams. They came in talking about streamlining onboarding, but what they really wanted was to understand why some managers got people functioning in weeks while others took months. Once we reframed it as capturing what the best managers knew, not just documenting the steps, everything opened up.
That's not a process problem. That's a knowledge problem.
The real opportunity isn't efficiency. It's intelligence. It's capturing the knowledge that lives in people's heads, the unstated information and learned experience that makes your best people your best people, and making it accessible, repeatable, and buildable. That's a different conversation. And when someone makes that shift, everything opens up.
The Misconceptions We Keep Correcting
For a lot of people, AI means ChatGPT. That's the frame. And it's not wrong, exactly, but it's incomplete. It's like thinking the internet is email.
The pace of change here is hard to overstate. What was true six months ago isn't necessarily true now. And yet I still meet leaders who believe they can wait it out. Let things settle. Find their footing once it's all more certain.
I understand the instinct. But I think it's a mistake. The companies that are going to lead in this next era aren't waiting for certainty. They're building the capacity to learn and adapt. That's the real skill now.
If you're not sure where to start, try this: ask what your best people know that isn't written down anywhere. That's usually where the real leverage is.
Maturity Isn't Binary
One thing I've stopped doing is thinking about companies as "ready" or "not ready." It's more nuanced than that. Maturity shows up differently across departments, teams, and even individuals. You might have a CFO who's thinking five years ahead and a sales team still asking if AI can write their emails. Both realities often exist in the same company.
What I've noticed is that the most productive conversations happen when someone stops asking "what have you done in my industry?" and starts asking "what have you learned?" That shift, from validation-seeking to curiosity, is the tell. It means they're ready to think, not just shop.
The mature buyers aren't vetting tools as magic bullets anymore. They're asking how to refine processes, how to think differently. They want to know what we've learned, how they can work differently. You can tell when someone shifts from defending what they already know to exploring what they could create.
That's when deeper conversations about knowledge, customer experience, and business transformation become possible.
Where We Are Now
If I'm being honest, I expected the market to be further along by now. I thought more leaders would understand why agent architectures matter, why infrastructure decisions are crucial. I thought we'd have moved past transactional AI usage, people using ChatGPT like a better search engine, and started building real context. But most organizations are still early. Very early.
In practice, that looks like a few tools scattered across teams, some failed attempts at AI initiatives, IT treating it as a problem to solve rather than a capability to build. There's a lot of forward momentum followed by backtracking as new information arrives. It's flux.
Early on, most of our conversations were about education. What is this thing? How does it work? Is it safe? Now the questions have evolved. People want to know what it can do. They're still not designing solutions, not most of them, but they're leaning in. The market is maturing, unevenly and sometimes painfully, but it's moving.
We've been moving too. Refining what we've learned. Building around it.
What's Coming
What we're building isn't based on assumptions, trends, or what the headlines say AI should be. It's based on what we've heard. Directly. Repeatedly. From people trying to figure this out in real time.
We've done the work. And what's coming reflects that.
More soon.
Get more from Impact’s AI team in the webinar, Lessons Learned from Incorporating AI into Business Processes.
