Transformation Isn’t a Tech or Strategy Problem. It’s a System Issue.

What I’ve learned from $10B+ in transformation projects: if you ignore the system, the system will ignore your strategy.

I wanted to come back on some thoughts around project Tamkeen, which was a collaboration between University of Warwick - Warwick Business School and Emirates. Beyond the experience (covered here), it reminded me of how different generations and companies approach their transformation initiatives, including AI deployment.

The Pattern I Keep Noticing

95% of the time, the focus will be on the new capabilities, the product features, and technical integration.

That's of course needed, but where we need to spend more of our time is actually on execution risk, rollout sequencing, and adoption failure modes.

Your team needs to be able to easily answer: "How can we ensure adoption momentum from day-1?” …because risk aversion tends to get stronger as you move up the leadership ladder, and initiative fatigue is visible across thousands of organizational front lines.

My roll out playbook (whether AI or not)

Instead of starting with "How fast can we get this across?", begin with a few questions:

• Where does it make strategic sense to start? Not all departments/regions are equally ready.

• How should we sequence the rollout across the organization? Build your formula to capture: infrastructure readiness × strategic impact × risk/cost x ...

• For every BU, ask: “What can kill adoption? Change fatigue, cultural resistance, complexity overload… find the risks.

• How do you measure real success? Go beyond deployment metrics and get to actual behaviour changes at employee, customer, and group performance knock-on level.

A framework that works

Whatever shiny object is calling you, ensure you have mapped out your "Capability vs. Opportunity Matrix".

Plot initiatives across two (or more) dimensions, and if data allows, as part of an equation that also considers your operating model and financial scenarios, - but at a minimum, capture:

• Ease to implement (tech readiness, regulatory complexity, CapEx)

• Strategic impact (savings potential, innovation lift, employee experience)

That should already give you a clearer prioritization: Quick Wins => Scale What Works => Then Moonshots. And it also shows what not to focus on, because knowing what NOT to do, can often be more valuable than what TO do.

Once you have focus points, build out your  "Path Forward" phases:

1. Discovery: User journey mapping + stakeholder alignment

2. Prototype: Low-code testing with real users

3. Pilot: Measured rollout with defined success criteria

4. Integrate: Scale with continuous feedback loops

Start where value meets readiness. Sequence what matters and don’t confuse visibility with viability.

Experience vs. Innovation Divide

Experience means some “battle scars” from previous transformations – or as in my case, prematurely grey hair. Here, cross-functional/silo experience is critical to understand where initiatives typically crash and burn, but also where to apply oil in the internal machine. And since some internal initiatives will also impact your customers and stakeholders, experience helps understand which industry and market counterparties can help with adoption momentum.

Complement that by seeking viewpoints from younger strategists and professionals. These bring fresh perspectives on what's possible, whether AI or other - they're not constrained by "how things have always been done." And even if some suggestions can appear completely unhinged from market realities – they should be listened to, because markets are dynamic.

When I speak about having an inclusive environment, it’s about allowing these different perspectives to be considered and stacked. In the case of AI, it means an ambitious AI vision grounded in realistic execution planning, tech stacking, and market pragmatism.

So about that AI ROI

Research¹ highlights the benefits of AI: better forecasting, faster decision making, improved cycle time, and how companies adopting GenAI are redesigning workflows end-to-end, allowing cost cuttings across the org chart.

But here's what your boardroom presentation probably does not say upfront: your transformation will likely fail if your current execution rate is already poor.

Because AI or not, the traditional transformation failure rates remain stubbornly high. McKinsey research shows 70% of business transformations still fail, with digital transformations faring even worse.

So mind the gap – because it's not technical. It's political.

In some shape or form, I’ve been involved in over $10 billion worth of capital expenditure rollouts. And in 99% of large organizations, the pattern is the same:

Everyone nods in agreement during the meeting session, but the informal conversations that follow, whether on WhatsApp, Signal, or behind closed doors, often reveal the real probability of success.

Even when an AI initiative, or other, makes perfect sense for the company, it will face roadblocks you never see coming:

• The CFO who publicly supports it, but privately questions the ROI and timeline, adding to his frustrations managing the balance sheet and the CEO's vision.

• The COO who fears it threatens their operational control over parts of the supply chain, or how their team is benchmarked for variable compensation.

• The CHRO concerned about workforce displacement messaging, succession, and employee engagement.

• Regional heads who see it as just another "HQ initiative", where waiting for it to fizzle out is a viable strategy.

The gap isn't "Can AI do this?"

It's "Will the CEO's peers on the ExCo actually execute this when the pressure mounts?"

What This Means for Your AI Strategy (because you still need one)

Before you build your AI solution, map the real constraints.

Technical readiness - everyone focuses here:

• Digital infrastructure, integration complexity, data quality, etc..

Political readiness - where most projects actually die:

• Which ExCo members have hidden reservations?

• Who loses (or gains) power/budget/headcount if this succeeds?

• What happens when quarterly results pressure mounts and this becomes the "expensive experiment"?

• Whose seat is already most threatened? (there is always one)

Execution muscle - the make-or-break factor:

• Can you actually deliver change management at scale?

• Do you have credible champions who survived previous transformations?

• When push comes to shove, will leadership stay committed through the messy middle when some engine troubles are felt?

• Which division, BU, etc., needs additional support?

...and success metrics that matter:

• Not just deployment milestones, but real behavior change.

• Not just efficiency gains, but sustained adoption, 3, 6, 12 months later and beyond.

• Not just the CEO's enthusiasm, but the frontline, kitchen - and yes, even casual toilet conversations.

Because the most sophisticated AI in the world is worthless if the organization's antibodies reject it six months post-launch.

After $10B+ in CapEx across large-scale transformation efforts, and watching brilliant strategies ground themselves on the rocks of organizational politics, I’ve learned this: The kitchen, WhatsApp, and Signal conversations are your real business case. Everything else is just PowerPoint. People will remember and discuss the mess, so it helps to get it right from the start.

What's your experience? Have you seen AI initiatives that made perfect strategic sense but died due to internal resistance? And for senior leaders - how do you navigate the gap between public support and private skepticism in your transformation efforts?

¹ Sources: McKinsey & Co, BCG procurement transformation studies

#AIStrategy #Transformation #ChangeManagement #Leadership #Innovation