The Pilot Worked. The Rollout Didn’t. Now What?
How to figure out what is blocking you and get moving with AI transformation
Hi there,
It’s been a massive few weeks in the AI world - every major AI lab has released a new model, and there is a lot of new functionality. In this article I used Google’s Nano Banana Pro to make the infographics - what do you think?
AI models keep getting better every few weeks, but there are a lot of leaders and workers wondering why their AI pilots work out but as soon as their organisation tries to do a production rollout of an AI solution, it all falls down.
What’s Going Wrong With Blocked AI Projects
Your pilot worked. The business case was solid. The board said yes.
And now, six months later, nothing is moving. Usage is flat. The teams who were supposed to adopt it are finding workarounds. Your middle managers nod in meetings and then do nothing. The executive who sponsored the whole thing is starting to ask pointed questions.
You have an initiative that is stuck. And you are starting to suspect that the problem is not the technology.
You are right. It is not the code; it is the culture.
BCG’s 2024 research found that 74% of companies struggle to get value from their AI investments at scale. The pattern is consistent: the technology works, but the organisation does not change. The tools sit there. The dashboards go unread. The promised productivity gains never arrive.
This playbook covers how to diagnose what is really blocking you and how to get things moving again in the next 90 days.
Step 1: Figure out what is actually blocking you
Most leaders assume they have a technology problem or a training problem. They rarely do. They have a people problem, and usually a specific one they have not yet identified.
Resistance is not a single thing. Different groups resist for different reasons.
Frontline staff are worried about their jobs. A 2025 survey found that 89% of workers have concerns about job security related to automation. If your frontline teams are dragging their feet, fear is almost certainly part of the picture.
Middle managers are facing an identity crisis. Only 34% of managers feel equipped to support new technology rollouts, according to BearingPoint’s 2025 research. They are being asked to champion something that might eliminate their own roles. Their response is often to slow things down while they figure out what this means for them.
Specialists are protecting their expertise. Lawyers, accountants, clinicians and engineers often resist tools that seem to commoditise what they do. They are not worried about the technology being wrong. They are worried about what it means for their standing.
Risk and compliance teams are doing their job. If your legal or risk functions are blocking progress, it is often because they see genuine problems the project team has not addressed. Accountability gaps, bias risks and data governance issues are real concerns that need real answers.
There is another factor that compounds all of this: your people may be exhausted. Gartner’s research found that employees’ capacity to cope with change has dropped to half of what it was before the pandemic. If your organisation has a history of starting things and not finishing them, that history is working against you right now.
The move: Write down the three to five groups whose adoption matters most. For each one, identify what you believe is driving their resistance. Then go and check. Do not ask leading questions like “Do you like the new tool?” You will get false positives. Instead, ask: “If you had to use this tool for every task starting tomorrow, what part of your day would break?” That question surfaces real objections.
Step 2: Get your middle managers on side
Harvard Business Review research found that 60% of change initiatives fail at the middle management level. These are the people who translate strategy into action, who allocate their team’s time, who signal through their own behaviour what actually matters.
If your middle managers are not genuinely behind your initiative (not just saying the right things in meetings, but actually prioritising it, using it themselves, coaching their teams) you will not succeed. It does not matter how good your technology is.
McKinsey’s research shows that when middle managers are properly engaged, teams are five times more likely to achieve successful outcomes. The difference is not marginal.
Here is what engaging them actually requires.
Include them in decisions, not just announcements. If the first time they heard about this initiative was when it was already decided, you have a trust deficit. Go back and involve them in how it gets implemented, even if the “what” is already set.
Address their personal uncertainty directly. Do not pretend their roles are not affected. Be honest about what is changing and what is not. If you do not know yet, say so, but explain how and when you will figure it out.
Give them what they need to lead the change. Two-thirds of managers do not feel equipped to support new technology rollouts. Training, resources, talking points and permission to experiment and fail are not optional.
Subtract before you add. This is the most important point. Middle managers are already stretched. You cannot add responsibilities without taking something else off their plate.
The move: Explicitly de-prioritise one existing report, meeting series or initiative to “buy” your middle managers’ capacity for this. Make the trade visible. If you cannot identify anything to subtract, that tells you something important about whether this initiative can succeed right now.
Step 3: Create a visible win in 90 days
Research on change initiatives consistently finds that early wins are essential for maintaining momentum. Kotter’s studies found that without tangible improvements within 6 to 18 months, old habits return and initiatives lose steam.
But not just any win will do. A useful early win has three qualities: it is visible (enough people can see it), it is unambiguous (no reasonable argument about whether it worked), and it is clearly connected to the initiative (not something that would have happened anyway).
Manufactured wins do not help. If people sense you are dressing up a small improvement to look like a bigger deal than it is, you lose credibility.
For a stalled initiative, this usually means narrowing your focus. Rather than trying to get everyone to adopt everything, pick one team, one use case, one workflow and make that work properly. Get a group of people genuinely using the tool in a way that makes their work better, and make sure the rest of the organisation can see it.
The move: Identify your best candidate for an early win. Which team is most receptive? Which use case is most likely to deliver an improvement people will actually notice? Who has enough credibility that others will pay attention when they succeed? Focus your energy there for the next 90 days. Get that working before trying to expand.
Step 4: Fix your communication
Prosci’s research shows that organisations are three times more likely to succeed when employees genuinely buy in, and that clear communication doubles success rates. But more than half of senior HR leaders say their current communication is failing to engage employees.
Most communication about technology initiatives falls into one of two failure modes. The first is too abstract: vision statements and strategy decks that do not answer “what does this mean for me?” The second is too technical: feature lists and implementation details that do not answer “why should I care?”
The research identifies three types of resistance that communication needs to address. Most organisations only address the first.
Logical resistance requires information: data, evidence, clear explanation of the rationale. This is what most corporate communication focuses on.
Psychological resistance requires reassurance: acknowledgment of concerns, honesty about uncertainty, support for the transition. This is often missing entirely.
Sociological resistance requires working with informal leaders and peer groups to shift norms. This is rarely even recognised.
The messenger matters as much as the message. Business rationale should come from senior leaders. Personal impact should come from direct supervisors. Prosci found that 58% of employees prefer to hear about how change affects them from their immediate manager.
The move: If your initiative has been communicated primarily through all-hands announcements and leadership emails, you are missing the most important channel. Equip your direct managers to have one-on-one conversations with their teams about what this actually means for them, and make sure those conversations address psychological and sociological resistance, not just the logical case.
What to do on Monday
This week: Write down the 3 to 5 groups whose adoption matters most. For each, note what you believe is driving their resistance.
This week: Talk to at least two people from each group. Ask: “If you had to use this tool for every task starting tomorrow, what part of your day would break?”
Within two weeks: Have honest conversations with your key middle managers. Find out what they need, and what you can take off their plate to make room.
Within 30 days: Identify your best candidate for an early win. One team, one use case, one workflow. Narrow your focus there.
Within 90 days: Deliver that visible win, and make sure the rest of the organisation sees it.
A note on getting help
Everything in this playbook is designed to be actionable without outside assistance. If you work through these steps honestly, you will have a much clearer picture of what is blocking your initiative and what it would take to move it forward.
If you work through the Monday list and find you cannot answer the questions about resistance or capacity, or if the conversations with middle managers surface problems bigger than you expected, that is worth paying attention to.
I offer a 60-minute working session focused specifically on stalled initiatives. We diagnose where you are actually stuck, identify the few moves that would shift adoption in the next quarter, and you leave with a clear picture of what is really going on.
You can book a paid call here. But the playbook above will get you a long way on your own.
Regards,
Brennan




Spot on. This is so true! As a techer I see it all the time with new ed-tech too. The culture shift is definitely the hardest part, not the technology itself.
Great advice! I've seen this happen way too often
One more thing: get the c level on board of you can. Almost nothing is a bigger predictor of AI initiative success than support from the top. There's only so much goodwill you can wiggle from the ground up