Volume 1 | Issue 1
October 2024
What I Keep Hearing From Leaders About AI Transformation (And What the Data Says About It)
Self-limiting beliefs
and self-imposed
mental barriers might
be holding you back
Of late I’ve been having a lot of conversations with leaders driving transformation in organisations and have picked up some very interesting threads. Not the polished stories — the real ones, usually about twenty minutes in when we get past the official version.
Almost everyone is under pressure to move on AI. And almost everyone, privately, is unsure whether what they’re doing is actually going to work.
That admission is more common than you’d think. And looking at the data, it’s also more justified than most organisations are publicly acknowledging. A 2025 MIT study of 300 public AI deployments found that around 95% of enterprise AI pilots fail to create measurable impact on the bottom line. S&P Global’s research puts the abandonment rate for AI initiatives at 42% of companies in 2025 — up sharply from just 17% the year before. McKinsey reports that while 88% of organisations now use AI in at least one function, only 6% qualify as genuine high performers seeing significant financial returns.
That gap between activity and impact is, quietly, what many of the leaders I talk to are dealing with right now.
The part nobody talks about: we don't always know our own workflows
Here’s something that comes up again and again. When teams sit down to figure out where AI can genuinely help, they often discover they don’t have as clear a picture of their own processes as they thought. Not because anyone has been asleep at the wheel — but because when you’re running hard, there’s no time to step back and trace how work actually moves. The workarounds, the informal handoffs, the gaps between how things are supposed to happen and how they actually do.
McKinsey’s research is pretty direct on this point: among all the factors they tested for driving real AI impact, redesigning workflows had the single strongest correlation with meaningful business results. Their 2025 report found that high-performing organisations are almost three times more likely to have fundamentally reworked their processes — rather than simply layering AI on top of existing ones. Yet only about 21% of organisations using AI have done this so far.
This is where an outside perspective becomes genuinely valuable — and I don’t mean a big consultancy arriving with a pre-built framework. I mean a small, specialist project team that comes in without an agenda, takes the time to understand your specific context, and helps you map things honestly. The value isn’t that they know more than your team. It’s that they’re objective in a way that’s almost impossible to be from the inside.
That map — built collaboratively, with the internal people and the project team who are there to augment your internal team — tends to surface the right starting points. Which is a lot better than guessing.
..high-performing organisations are almost three times more likely to have fundamentally reworked their processes — rather than simply layering AI on top of existing ones. Yet only about 21% of organisations using AI have done this so far.
Co-creating the plan rather than receiving one
Something I’ve noticed: the organisations where transformation is actually taking hold are the ones where the plan feels owned internally, not handed over. There’s a difference in ownership and engagement between a roadmap that gets built with a team and one that gets delivered to them, and one can feel teh difference immediately.
What seems to work is having external specialists and internal people in the same room, combining the objectivity and cross-sector experience of one with the institutional knowledge and credibility of the other. MIT research reinforces this too — organisations that partnered with specialist external teams to implement AI succeeded roughly twice as often as those who tried to build everything in-house.
The same applies to tool selection, which can feel overwhelming right now given how much is out there. The filter that tends to cut through is whether it fits how the team actually works, integrates with what’s already in place, and holds up against real security and compliance requirements — not just in theory but in practice.
There's a difference in ownership and engagement between a roadmap that gets built with a team and one that gets delivered to them, and one can feel the difference immediately.
Being in the work together, not just advising on it
The pilots that lead somewhere tend to have something in common: the external team is actually in it with you. Co-executing, not observing. Helping build guardrails as things come up, not writing recommendations about them afterwards. Working out what success actually looks like — in terms that matter to the business — before anything gets measured, not after.
This matters more than it might seem. Informatica’s 2025 research found that data quality and readiness is the number one obstacle cited in failed AI projects, followed closely by lack of technical maturity and skills gaps. These aren’t problems you can anticipate fully from the outside looking in — they surface during execution, which is exactly why having specialist support through the pilot phase, not just in the planning phase, makes such a difference.
There’s also something about the quality of feedback that changes when everyone involved has a stake in getting it right. You get more honesty about what’s not working, earlier, which is exactly when it’s still useful.
..organisations that partnered with specialist external teams to implement AI succeeded roughly twice as often as those who tried to build everything in-house.
The people thing is harder than most plans account for
This is the part of the conversation that tends to get quiet. The technology is usually fine. The harder question is what’s happening with the team.
A 2024 EY survey found that 75% of employees worry AI could eliminate their jobs, with 65% specifically concerned about their own roles. NTT Data’s research cites that 45% of workers are already experiencing burnout from the pace of organisational change — and Prosci reports that 75% of organisations are at or past the point of change saturation. Layer an AI transformation on top of that, and the conditions for resistance are already in place before you’ve even announced anything.
My sense is that the leaders who are navigating this well aren’t necessarily the ones with the most sophisticated tools and change management frameworks. They are the ones who communicated early with their teams before they had all the answers, who brought people into the process from the get-go to actually influence the direction, and who identified trusted internal voices in each team to carry the message forward. McKinsey’s own research is blunt about where the real barrier sits: employees are ready for AI — it’s leadership that’s lagging. Research shows that employees are three times more likely to already be using AI than their C-suite leaders estimate.
..communicating with genuine transparency throughout. Not a single announcement — an ongoing narrative that's been thought through in advance, adapted for different parts of the organisation, and held to even when the answers aren't fully formed yet.
Where that leaves us
There isn’t a fixed formula in play. Every organisation’s starting point is different, and the honest answer is that there isn’t a clean playbook that works everywhere.
What does seem consistent is that the teams making real progress are the ones willing to get clear on where they actually are before deciding where they’re going — often with specialist help — and then building the plan with their people rather than for them. They execute alongside partners rather than just being at the receiving end of their advice. They set benchmarks collaboratively, build guardrails to ensure process sanctity, and stay honest about what the data is telling them.
Of course there’s pressure to move fast. But most of the leaders I respect right now are doing three things quietly and consistently: planning before they move, bringing in short term teams to augment the change effort and communicating with genuine transparency throughout. Not a single announcement — an ongoing narrative that’s been thought through in advance, adapted for different parts of the organisation, and held to even when the answers aren’t fully formed yet. From what I can see, that combination — deliberate co-created planning, executing with practical hands-on external help and honest, consistent communication — is what’s actually making the difference.
Sources
- MIT NANDA Initiative — The GenAI Divide: State of AI in Business 2025 (Fortune, August 2025): https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/
- S&P Global Market Intelligence — Enterprise AI Abandonment Survey 2025 (WorkOS, July 2025): https://workos.com/blog/why-most-enterprise-ai-projects-fail-patterns-that-work
- McKinsey & Company — The State of AI 2025: How Organizations Are Rewiring to Capture Value: https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai
- McKinsey & Company — Superagency in the Workplace: Empowering People to Unlock AI’s Full Potential at Work (January 2025): https://www.mckinsey.com/capabilities/tech-and-ai/our-insights/superagency-in-the-workplace-empowering-people-to-unlock-ais-full-potential-at-work
- Informatica — CDO Insights 2025: Top Obstacles to AI Success (via WorkOS): https://workos.com/blog/why-most-enterprise-ai-projects-fail-patterns-that-work
- EY — 2024 Employee AI Anxiety Survey (via Cybersecurity Intelligence): https://www.cybersecurityintelligence.com/blog/employee-resistance-to-ai-adoption-8641.html
- NTT DATA — GenAI Deployment and Change Fatigue Research 2024: https://www.nttdata.com/global/en/insights/focus/2024/between-70-85p-of-genai-deployment-efforts-are-failing
- Prosci — Best Practices in Change Management, 12th Edition: https://www.prosci.com/blog/change-management-trends-2024-and-beyond
#StartWithI
#TransformationThatLasts
#DrivingEffectiveLeadershi
Anu D’Souza runs Bricoleur Consulting, a leadership and transformation services advisory. A thought leader on AI era leadership and CX/ EX alignment, Anu has spent many years with companies like Unilever, Ogilvy and BBDO and has lived and worked in multiple cultures running teams across borders. Anu is also the author of ALIGNED Why CEOs need Company Brand Alignment in the Age of a Questioning Workforce. You can reach her on anu@bricoleurconsulting.com or book a call here.