Your AI Adoption Strategy.

Ask an AI vendor how to get your people actually using the tools you've bought, and you'll get a training plan and a usage dashboard. Ask a change consultant and you'll get a stakeholder map and a twelve-week comms cascade. Both sound sensible. But sadly both are answering the wrong question.

Because here's the question that really matters:

"if the technology already works, why is hardly anyone using it?"

BCG has a number for this, and it should be printed on the wall of every transformation office. According to their research, when it comes to getting value from AI, here's the breakdown:

• Only 10% of the work is the algorithm
• 20% is the technology
• And a full 70% is people and process

So I'd argue that your AI adoption strategy isn't really a technology strategy at all. It's a behaviour-change strategy wearing a technology badge. And the discipline that exists to change how thousands of people behave, at scale, on purpose? That's marketing. You already pay for it. You've just never pointed it at your own staff.

Why most AI adoption strategies get it backwards.

Most organisations run their AI strategy upside down. They pour months into picking the model, wiring up the data, and arguing about which platform wins. That's the 10% and the 20%. Then they "launch" to their people with a couple of all-staff emails and a new training pathway invite, and call the 70% done.

Seriously, observe where the money is spent and what the meetings focus on. Model selection gets a steering group. Data integration gets a project plan. Security gets three workshops and a risk register. The actual humans who have to change how they work get a launch email and a link to a portal they'll open once. The 70% gets the scraps, and then everyone acts surprised when the numbers don't move.

This problem is already revealing itself in the numbers. MIT found that 95% of AI pilots deliver no measurable impact on the bottom line. And by late 2025, 42% of companies had walked away from most of their AI initiatives, up from 17% a year before. The tools didn't fail. The rollout did.

If you want to explore that more, I went deep on why that happens in a separate piece on AI adoption challenges. The nub of it is this: Canva gave 5,000 staff a week off work to do nothing but learn AI, and their own Chief Customer Officer concluded the blocker wasn't the technology. It was the people.

The lesson holds for everyone: you can buy the cleverest tool on the market and watch it gather dust if nobody's given people a reason to care.

Your AI adoption strategy is really a behaviour-change strategy.

Let's go back to BCG's 70%. If most of the value lives in people and process, then "adoption" is just a polite word for getting humans to change what they do every day. That's hard. Like, really, really hard. It's also a solved problem, because one single industry has spent a century doing exactly that: marketing.

Marketing doesn't train people, and it certainly doesn't mandate. It makes people want the thing. It works out who they are, what they care about, what's stopping them, and what would tip them over the edge. Then it says it again, and again, until the new behaviour sticks.

This is why the training-first instinct is so seductive and so wrong. Training feels like progress. It's tidy, it's measurable, you can put completion rates in a board pack.

But a completion rate isn't a behaviour. Teaching someone how to write a prompt does nothing with the problem of whether they want to use the prompt, whether they trust the output, or whether they can see the point for their own tasks. You can run every workshop going and still watch usage flatline, because you answered a question nobody was asking.

You've already got proof this works inside your own walls. Your people aren't anti-AI. In fact, 78% of them are bringing their own AI tools to work, unprompted, on their own time. They'll adopt AI enthusiastically. They just won't adopt the one you chose, because nobody sold it to them the way ChatGPT sold itself.

One caveat, and it's a big one. If your people are already running on empty, no amount of clever messaging will land. Change fatigue is real, and AI fatigue is its newest flavour. A strategy that ignores how knackered your workforce is will stall before it starts. Account for it now, or pay for it later.

The AI equation no one is solving.

Let's do a little experiment and try to get the total value of two different numbers. The first is what your organisation spends marketing to customers every year. The brand work, the demand generation, the agencies, the team, the media, the PR. It's a serious figure. The second is what you have spent or intend to spend on marketing AI to your own employees. For most organisations, that number is zero. Not "small". Zero.

So, to be clear, your org will happily spend a fortune persuading strangers to buy something once, and nothing at all persuading the people you already employ to use a tool you've already paid for, that would make them measurably better at their jobs. Said out loud, it's ridiculous and verging on bizarre.

That gap is the whole advantage. You don't need a new budget line and you don't need to hire some four-in-one-unicorn-human who's a technologist, a trainer, a change manager and a marketer all at once. Instead you just need to take a sliver of the thinking your marketing function already does brilliantly, and point it inward for once.

The AI adoption framework hiding in your marketing budget.

Whilst this might be quite frustrating in some ways, the truth is that the capability you need to tackle that 70% is already in your building. You're already paying for it. It's called your marketing team.

Think about what a marketing team does for customers every single week. They segment an audience instead of blasting everyone the same message. They build a value proposition. They test what lands and bin what doesn't. They launch things properly, with a build-up and a follow-through, not one email and a prayer. And they measure what people actually do, not whether someone opened a link.

Now look at what a decent AI adoption framework ACTUALLY needs:

• Know your audience. Finance, sales, operations and the front line fear different things and want different things. One message fits nobody.
• Answer "what's in it for me" for each of them, in their language, about their actual job.
• Launch it like a product, not an announcement. Anticipation first, then reinforcement, over weeks.
• Build visible champions and local wins. People copy their peers, not a memo from on high.
• Measure behaviour and sentiment, not licence logins. Logins tell you who opened it. They tell you nothing about who changed how they work.

So, you don't tell the finance team "we've rolled out AI". You show them how it kills three hours of manual reconciliation on a Friday afternoon, in their language, with a colleague they recognise saying it out loud. You don't tell sales "usage is mandatory". You show them the rep who clawed back two hours a week and spent it closing new deals (and making more commission as a result 🤑).

Specific, local, and believable. That's marketing, and it's the opposite of an all-staff email.

This is not some dull, humdrum change-management framework. It's a marketing plan. The unfair advantage isn't that you need to go and buy a new capability. It's that you already own one, and you've spent years aiming it exclusively at people who don't work for you.

If you want the full version, the five barriers to AI adoption and how to overcome each one, I've covered it all in this new guide.

What a marketing-led adoption strategy actually delivers.

This isn't theory. It works. For example, when we ran an AI and digital-skills rollout at Capgemini like a marketing launch instead of a training exercise, 82% of a 340,000-strong workforce engaged with it. Another of our clients went from 5% platform adoption to 23% in thirty days. No new software. No bigger licence. Just the existing tool, marketed properly to the people who were meant to use it.

The difference wasn't the budget or the technology. It was treating employees like an audience we had to win over, not users we could instruct. Same tool, different strategy, a completely different result. And it wasn't a fluke: the Capgemini approach won Awareness Campaign of the Year at the 2024 British Training Awards. Different technology, different tools, but the same principle applied every time.

BCG's latest thinking lands in the same spot. They now call AI transformation a workforce transformation, not a technology project. And I'd agree with that. To me, the companies pulling ahead aren't the ones with the best model, (because everyone can buy the same models and the playing field is equalled because of that). They're the ones whose people actually use them.

The unfair advantage is only unfair if you use it.

So here's where it leaves you.

Your competitors have the same tools you do. The same ChatGPT, the same Copilot, the same shiny platform the vendor demoed last quarter. The technology isn't anyone's advantage any more. How well your people adopt it is the only thing left to win on.

And the cheapest, fastest, most proven way to win it is sitting in your marketing budget right now, pointed in the wrong direction. Turn it around.

Here at MAAS, we don't build the tech. We make people use it. If that's the bit you're missing, let's talk. Or grab the guide and start clearing those five barriers tonight.