813 | The BIGGEST AI Mistake Most Companies Can Avoid in 2026 | Dajana Achelpohl

AI is in your company, whether you’ve officially launched it or not. People are using AI, but not the AI you want them to use. That’s Dajana Achelpohl, AI adoption strategist and founder of AI Changemaker. With nearly two decades leading global teams at Google and PayPal, she’s perfected frameworks that now help non-technical teams across agencies and enterprises confidently adopt AI in their day-to-day work. So how does she do it? I ask leadership, why are you rolling out AI? What is the business problem you’re trying to solve? And they do not have an answer. I use a framework that I call AI-building blocks. In this episode, you’ll learn why 70% of AI rollouts fail quietly.

I keep hearing, we bought the tool, but we’re not seeing results. Every time it comes down to how to spot and fix invisible resistance that’s sabotaging your team’s results. People were just worried about their jobs and that AI would ultimately take over. So they were actively sabotaging AI and making sure that the results were not good. And her hands-on blueprint for building an AI-ready team that your agency can get behind. We’re not coming to the group with a ready-made, this is your AI team charger and this is how we’re going to use it. We actually work through it and create it together. Companies that we’ve looked at, it feels like everybody’s using AI in one way or another.

But a lot of people and organizations aren’t using it effectively in. And what are some of the invisible blockers that you see to AI adoption in the organizations that you work with? So I worked in Google and PayPal for many, many years. What I see is a lot of hype around AI and less actual results. And I think there is a lot of pressure on doing AI because everybody’s doing it. So that goes right from the top down to the people actually doing the work, right? We are supposed to do AI. So people sometimes find it hard to admit that it’s actually not working or that it’s hard. So I think there’s a lot of people saying that they’re doing AI.

But if you dig a bit deeper, there is still a lot to learn and a lot more to do for most people. Leadership is saying we need to do AI. Everybody says yes, because that’s what’s expected. But what people are really doing is just doing their work the way that they’ve always done it because they don’t really see how AI is better than the way that they’ve done things before. I remember working with a team where we figured out that months after they had actually rolled out their AI tool, had launched this to the whole team, had done some training, only 15% of the people had actually opened, even just opened the AI tool after. So I think there’s a real issue with that silent non-use where people think we’ve bought the tool, all is good, but in reality teams are not using them.

So that would be one big thing. But silently not using it is one thing. Then another thing that I see is people actively sabotaging AI. So I’ve come across teams that wanted to roll out AI that had all the tools where people were just worried about their jobs and that AI would ultimately take over. So they were actively sabotaging AI and making sure that the results were not good. So I think you have these two extreme ends and then there’s a lot of variation in between. Another thing that I think is not often talked about or not enough talked about is shadow AI, where people are using AI but not the AI you want them to use.

That happens both in organizations that are not quite ready for AI yet and that say we actually don’t allow AI use on companies that have rolled out AI but very restricted tools. And then you see the scenario where people are bringing their own AI on their phones or in whichever way. And of course they are, why wouldn’t they, right? AI can help you do things quicker, but it’s a huge risk for an organization because you don’t know what people are using the AI tools for. If they’re feeding sensitive data into it, it’s a real mind field. So to me what everybody should really do in their organizations is have an open and honest conversation about AI.

Because AI is in your company, whether you’ve officially launched it or not. Fears around AI are something that is so important to address. It’s like the elephant in the room in many organizations where I come to talk about AI. Because people are wondering and how would they not? Like some of the headlines you see in the press about people being let go because of AI. So I think it’s really important to understand why people are resisting AI. If it is because they don’t have the knowledge, etc. then you should absolutely help them gain that knowledge. If it is fear, I think it’s important to address that too. And to me where I make a lot of progress with people that have that fear is talking about the skills that AI can’t cover at the moment.

There’s so many very very human things when we talk about things like critical thinking, true creativity, more sort of differentiated approaches to things. Where AI is not doing a good job. So I’m always encouraging people to think about it that way. There might be things that which they are currently doing, which AI can do in the future. There’s certainly things which they are already doing, which they can do more of. Which AI will not be able to do in the near future. So I think it’s important to have that conversation. Let’s talk about the scenario where you’ve got a team or a large organization. And maybe the majority of the team are more non-technical people.

And so you’ve got a situation where the tools have been put in place. But when you look at it, you’ve got sort of a small number of people really adopting them. So I hear a lot, we bought the tool, but we’re not seeing the results. And then when you dig into it, it’s always the same thing. Teams don’t know where to start. They don’t know how to apply AI. Or they don’t know what success looks like. A lot of that always comes back to what business problems we’re actually trying to solve. And you’d be surprised, Jeff, how many companies I work with where I ask leadership, why are you rolling out AI? What is the business problem you’re trying to solve? And they do not have an answer.

They’re doing AI because they think they should be doing AI. So what I think is super useful then to bring the whole team together, especially if they are non-technical teams, is go through a bit of a process. I use a framework that I call AI building blocks. Join the insights, tap that subscribe button, and join a community of agency owners looking to scale smarter, not harder. But just just that, right? A number of building blocks that really enable everybody in the team to use AI successfully. And we start again at that problem. Why are we using AI in this organization? What do we want AI to achieve? What are we not using AI for? Which is very important. Then there is basic AI literacy where people need to understand how to use AI.

Prompt engineering was the big thing for a while. I’m not a big believer in crafting super complicated prompts to get results out of AI. But I think it’s still important people understand how to get good results out of AI. We follow that then with a look at how can we use AI safely? So what are we using AI for? What are we not using AI for? Where we’re often talking about sharing customer data, et cetera. So how do we deal with that? Then there’s a bit of a block around ethics, data security, et cetera. And then we end the whole session with thinking about how can we actually keep this going? AI is no one and done. You are going to have to talk about this constantly.

You’re going to have to continue to learn. You’re going to have to talk about your successes and your failures. So all of these things are necessary. So I find it quite helpful to work through this and workshop setting with teams. And I find that after we have an actual problem that they want to solve, we have an excited group. And that’s really when you see a good uptake and usage of whatever tool you have. It’s the small wins that really are going to gash your team more invested into AI. And I’m happy to share an example. So I worked with a team that spent about four hours each team member every Friday to create a report. We can end a week report that they were sharing with stakeholders.

Nobody liked doing this work. It was tedious. It was something that people saw as a necessary ego. We used AI for it. That team was using Microsoft for quite a bit. You could have done that with any other tool that is available. And we cut the time to 20 minutes. So now instead of spending four hours on a Friday, people are only spending 20 minutes every Friday on doing this report. And to me, those kinds of wins are what really makes a difference because now people are really seeing what’s in it for them with AI.

This is me not doing the kind of work that I do not want to do. And it’s like little light bulbs going off and then people go, oh, if this works, then this other thing could also work. So I find really showing these smaller wins makes all the difference. And then you have that activated and engaged workforce that wants to do more of this. Because nobody likes doing a report. Let’s face it, Jeff, I’m sure you also don’t.

I think AI is one of the things where you don’t get away with only talking about AI. I think AI is a prime example of something where you need to get your hands dirty and use AI. I still hear a lot of people talk about AI. And by the way they talk about AI, you can kind of tell that they’re probably never used AI. So if we can leave people with one thing, it is start using AI for something. And it doesn’t have to be anything groundbreaking. Start by planning your next holiday using AI. Have it do your shopping. Have it organize your kids after school activities. But whatever it is, because this is not going to go away, right? This is going to be part of all of our lives.

So I think getting acquainted with it and somewhat fluent is what we are going to have to do sooner or later. So you might as well start now. You know, you were talking a little bit about the dark usage of it inside of organizations.

So it must be equally challenging for more technical teams or technical organizations to put guardrails in place and make sure that the organization is using it the way that they want it to. I see with more technical teams where it’s less of a barrier to use it. Everybody is super keen to use it. You have to put those guardrails into place. And what I think is important to note with that, I think when we hear guardrails, we often think guardrails is what’s going to limit innovation. When there’s too many rules around things, we won’t actually get all the goodness and the innovation out of AI. I actually find that those guardrails are really given people the security to know what’s okay and what’s not okay.

So I’m a big believer in having a common understanding in any group of what we’re using AI for and what we’re not using AI for.

Most organizations will have some sort of policy in place, like an official AI use policy, but that often doesn’t really translate into anybody’s reality. So I think it’s a good idea to have that formalized. There is an exercise which I do, which I call an AI team charter, literally the group working through all of these points and then reaching a common agreement. What are we going to use AI for? What are the tools that we are using? What are the tools we are not using? And I usually get them to sign it at the end so that it’s a bit of a contraction and a commitment and that it gets regularly reviewed by the group. And I find with groups who are further along, either because they’re more technical or just love to experiment more, having that in place is also something that really helps to put some structure in some direction on the AI use.

I also think I usually include a part of it. How do we add new things to AI? If you are showing people that, them identifying areas where AI could help and suggesting that is something that is encouraged, something that you want them to do, then that is something that you’re going to see so much more. I usually find that everybody feels more part of the AI journey after you’ve done an exercise like that. And it’s not a scenario where you just have some very keen, very technical people owning AI for everyone, but that everybody has a sense of ownership of what’s going on with AI. And again, I think that’s all you want. You’re never going to have a successful AI adoption because the leadership is buying the tool and telling everybody now we’re using AI.

You’re only going to have successful AI adoption if a vast majority of your team wants to use AI and wants to make it work and if they are bringing you the best use cases for AI because they are the ones that are doing the work. They were no way to make a difference. What makes this different than another policy document? Because I’m just thinking in the back of my head, oh my God, we’ve got a lot of policies around here and they aren’t always followed very well. The way I approach this with teams is that we’re building it together. So we’re not coming to the group with a ready-made, this is your AI team charger and this is how we’re going to use it.

We actually work through it and create it together because there might be certain things which an organization must have. Then we respect that. But ultimately, this is something that the group has actually created themselves. So not something that’s come in top-down and that will then really help them feel that this is different than other policies. I also always recommend having very regular reviews and looking at this again and having somebody on the team who owns this and checks in. And I know that it still might be something that people don’t think about all the time, but at least they had a hand in creating it, which I think makes it huge. Things are changing so fast.

So maybe throughout the next year here, are you going to see the people that you work with really pick up on the adoption of AI? I think we’re in a massive hype curve still with AI at the moment. So while I do absolutely see adoption growing, I think we are also going to see the hype tapering off a bit. So I’m hoping to see a lot less hype and a lot more tangible results and people actually using AI and doing some of the things that AI can do well. As you know, there’s a lot of talk about agentic AI and agents at the moment. Ultimately, agentic AI requires you to have a clear workflow, right? And that’s again, something where people, I think, are under the illusion that you just tell AI to do something that’s just magically going to do all these multiple step things.

So any industry that has clearly defined workflows and a clearer process is definitely far ahead. I see I know most of your listeners work in the agency field. I think agencies, there is a lot that can be done for agencies and I see some really interesting approaches there. I do see quite a lot in the services industry. There are some other areas that I work with that are far, far behind. Any industries which have strong element of regulation are just not going to get there in time. So I’m working with a few healthcare companies. I’m working with a few financial companies and they are just at a completely different stage because they are even worried to start with any AI.

So I think that there’s a vast field but I think that a whole area of services, there’s a lot of potential there in the short term.