95%
That is the share of enterprise AI deployments that, according to Bain, McKinsey and recent cross-industry data cited by the CEO of Sam's Club in a recent Fortune commentary, fail to produce meaningful results. Ninety-five out of every hundred pilots, platforms, and "transformation programmes" that leaders proudly announce in investor updates. Quietly underperforming, quietly abandoned, quietly rewritten as lessons learned.
When a number that big sits in front of me, I do not reach for better tools, instead, I reach for a better question. Because if the technology is largely the same across winners and losers, then the difference has to live somewhere else. It has to live in how we lead.
That is the conclusion I reached after reading four very different pieces of work during the month of April: Fortune commentary piece from the Sam's Club CEO on purpose-driven AI adoption; a Pan-African Journal study on ethical leadership in the age of AI; a systematic review of digital transformation research in Business & Administrative Sciences; and a fresh structural equation modelling study of 441 managers in S-Curve companies. Four different voices, four different methods, one surprisingly consistent message.
Technology is the easy part. Leadership is the lever.
Let me walk you through what the research is actually saying, and what I would do differently every Monday morning if I were leading your team/department/organisation.
The finding that should stop every leader in their tracks
I've changed how I talk to clients. I ask a simple but powerful question: do digital leadership and AI capabilities directly drive sustainable competitive advantage? You might be thinking that if we don't jump onto AI capabilities, we'll be left behind. Someone else will eat the cheese. In my experience, digital transformation leadership has a strong, significant impact on sustainable competitive advantage. AI capabilities, on their own, does not. There's research to support it.
The AI capabilities that your board is asking you to invest in — the models, the platforms, the licences, the integrations — showed no direct effect on long-term competitive advantage in the study. They only mattered when they were channelled through organisational agility: the culture, leadership, and adaptive routines that turn a capability into an outcome. This is not a case against AI. It is a case against AI without leadership. It is also, if I am honest, the clearest piece of evidence I have seen for why so many AI programmes stall six months in. The tools arrive. The vendors present. The dashboards go live. And nothing changes, because nothing in the organisation was prepared to change.
What the rest of the research quietly confirms
Once you see this pattern, you start seeing it everywhere.
There are a few leadership failures we already know: Amazon & Workday's algorithmic biased recruiting tool, the COMPAS predictive policing system, IBM Watson for Oncology. The common factor in every case is not bad technology. It is absent leadership. Leaders who delegated ethical judgement to the system, who did not audit, who did not interrogate, who assumed the algorithm was neutral because the vendor said so.
Closer to home, SA just withdrew it's National AI Policy because they uncovered fictitious citations and references.
Sam's Club used AI-powered computer vision to cut member exit times by twenty-three percent across six hundred clubs. They eliminated over two hundred million repetitive inventory tasks, freeing their associates to build relationships with members. This sounds familiar. They gave field leaders enterprise-grade access to ChatGPT. This wasn't a technology decision. It was a principle decision: people-led and tech-powered. Purpose first, tools second. Always.
AI does not produce advantage. Leaders who use AI purposefully do.
What does this mean for how you invest?
Based on what I'm experiencing right now, most of the AI investment conversations I am invited into are being held backwards.
They start with the tool. Should we buy this copilot? Should we pilot this platform? Should we build this agent? Those are fine questions, eventually. But they are not the first question.
The first question is a leadership question.
First: what is the purpose this technology is serving?
Not the use case. The purpose. Borrow this framing: start with the biggest points of friction for your business, your people, and your customers. If you cannot name the friction in a single sentence, you are not ready to buy the tool. And if the purpose is "because everyone else is doing it," you have already joined the ninety-five percent.
Second: Is our organisation agile enough to absorb what this tool will do to us?
Be explicit. AI capabilities only create sustainable advantage when they move through the organisational system. That means your people can learn quickly, your teams can adapt quickly, your decisions can travel quickly, and your culture rewards value creation over perfection. If those conditions are not in place, the tool will either sit unused in experimentation paralysis or quietly cause harm. In both scenarios, the ROI conversation you are about to have with your CFO will be painful.
Third: Who owns the ethical oversight, by name?
This is the question worth keep returning to, and it is the one I find most leaders are least prepared for. Not "do we have an AI policy". Most organisations do, but "who in this room is personally accountable for fairness, transparency and explainability in this system?"
If the answer is "the vendor" or "IT" or a shared platitude, you have the same accountability gap that produced Amazon's hiring disaster and the COMPAS bias scandal. Youwouldn't outsource trust. In this same breath: ethical AI leadership cannot be outsourced. It has to sit with a named person, at a senior enough level to say no.
The human question, beneath...
There is one more thread I want to pull on before I close, because I think it is the one that matters most for South African and African organisations in particular.
AI amplifies whatever it is pointed at. Point it at a clear purpose and capable people, and you get the Sam's Club story — faster service, more fulfilling work, stronger member relationships. Point it at unclear intent and an underprepared workforce, and you get bias at scale, disengagement at scale, and a failed pilot at scale. You'll be faster at being slow.
The difference between those two outcomes is not the model. It is the knowledge, behaviour and attitude of the people who use it.
That is the quiet, unglamorous work that the headlines miss. It is not a model upgrade. It is a literacy upgrade. It is executives learning to ask sharper questions of their AI systems. It is HR and L&D leaders translating capability frameworks into practical skill building. It is managers learning to coach their teams through ambiguity instead of around it. It is employees feeling prepared rather than replaced. Agency, not anxiety.
Ok, you're in but where you would start?
If I were sitting across the table from you this week, here is the sequence I would recommend.
Begin with a diagnostic. Not a technology audit — a readiness audit. Where are your leaders on AI literacy? Where is your workforce? Where are the trust gaps, the skills gaps, and the purpose gaps that will silently strangle your next deployment? Most of the organisations I work with (until now) have never looked at those questions in one view before, and the conversation it unlocks tends to reshape the next eighteen months of investment.
Then build a leadership coalition, not a committee. Three to five senior leaders: operations, people, technology. Meet with discipline, own the purpose, the agility, and the ethics questions by name, and are willing to say no to shiny tools that do not serve those three priorities.
A quiet invitation
If this resonates, and you would like a calm, structured conversation about where your organisation actually sits on readiness, not a pitch, a conversation. I offer a free thirty-minute AI Readiness Diagnostic for executives and HR leaders. We walk through the three questions above, apply them to your context, and I send you a short written reflection afterwards. No slide deck required.
Write to us directly at devaan@cerebralconsulting.co.za.
For teams that are further along and want a structured starting point, our Strategy & Capability Blueprint equips executives, managers and employees with the shared language and decision frameworks to move purposefully rather than reactively. It is where most of our clients begin.
AI is not the advantage. The leaders who know how to use it are. The Age of AI will be led by organisations that choose purpose before promise, people before platforms, and clarity before speed. I would like to help you be one of them.