Why clarity, not code, is the key to AI success

Written By Stuart Maister

I was at an event last week in which the same solution came up again and again when six major business issues were discussed. These included the question of how to successfully use AI.

Those there dived in, workshop-style, to discuss how best to tackle these issues. What was striking to me that in each case the same or similar answer was top of the list.

Strategic clarity.

What was true in almost every case was that those within organisations were seeing confusion and conflicting priorities in almost every area of their operations. Actions differed from words. Behaviours didn’t match what was said. And the result was inaction or downright hostility to change, because in the face of this people act defensively and point fingers at everyone else. Trust breaks down.

Some I spoke with said that the culture in their organisation was destroyed by this uncertainty and lack of clarity, with people moving in different directions, doing what they felt was the right thing but not in a way that was consistent with others.

In one or two cases this was not true. There, real clarity meant team alignment, because everyone knows what’s happening and can point in the same direction. That works at a company-wide level, at the start of a change project or in the way you lead. It ticked every box at this event when it came to the top six business issues.

In the case of AI introduction, lack of clarity gets amplified because of the sheer power and speed that’s on offer. Get things wrong and mistakes are magnified.

AI treated as an efficiency lever, not a strategic amplifier

Multiple large studies show that a majority of AI initiatives never translate into sustained business value — not because the models are weak but because deployments are disconnected from an organisation’s strategic identity, ways of working, and no one explains why it might be in the interests of the people involved.

In the latest study, from MIT, the researchers emphasise that failures are rarely about the AI technology itself, but about organisational context, with lack of strategic alignment being top of the list. In short, even the most sophisticated AI cannot create value if it is not anchored in a clear strategy. This research suggests that, as a result, 95% of organisations introducing generative AI projects are getting zero return.

Too often AI is introduced as a point solution to remove friction or cut cost. That approach sometimes delivers tactical wins, but tactical wins are not a strategy.

What strategic clarity looks like (and why it matters)

A robust Strategic Narrative answers four interlocking questions:

·      why the organisation exists (purpose);

·      which clients it is built to serve (target segments);

·      why those clients choose it (differentiation);

·      and precisely what value it creates and how.

The how is critical: how are we special, different and great compared to others in the market? That’s usually where the magic lies in my experience. And then it must be articulated in a way that everyone understands and can use as a filter when making decisions.

Michael Porter’s classic definition of strategy—deliberately choosing a different set of activities to deliver unique value—underscores that strategic clarity is about positioning and focus, not operational effectiveness alone.

Aligning AI to that narrative ensures the technology amplifies distinct advantage rather than merely automating the same activity everyone else can buy.  If your focus is on technical delivery then you have nothing to differentiate yourself from others who can now also offer this.

Once again, this applies at a project as well as at a company-wide level. Having a clear Strategic Narrative around the project – or for the team affected by the project – means everyone can understand why and how this deployment makes sense in terms of the brand promise to the marketplace.

The Trust Triangle — the governance lens you can use immediately

The difference usually lies in the people, not the technology. Elsewhere I have written about how a focus on AI efficiency can destroy a company’s soul. I want to return to this idea here and spell out what to do about it.

The core thought is that an organisation has a soul – something intangible that is more than just culture, values or brand essence, and is the magic that makes the organisation a unique and integrated entity. So how do you preserve this in the way you introduce AI? The answer is to be consistent with who you say you are as an organisation.

The Trust Triangle in my co-authored book ‘Choose Trust’ is a good framework to use in planning the implementation of AI. The goal is to build and grow trust among those involved so that they can deal collaboratively with the challenges and opportunities presented.

There are three dimensions to trust as we defined it, and they are the drivers of this process.

  • Clarity (shared ambition and outcomes). Firstly, have the clear Strategic Narrative described above. Remember, this also defines your ‘how’, and that will be the recipe for the secret sauce that makes your organisation what it is. The AI is there to make this story even more true – even if it changes the ways in which you deliver some of your value. Next, openly translate the Strategic Narrative into explicit, agreed AI objectives: eg which customer decisions should improve, what metric will change, and what success looks like. According to McKinsey, the best performing AI adopters explicitly connect AI initiatives to strategy and measure them by strategic outcomes.

  • Character (the behaviours you commit to). These should be in line with your values and culture. Ideally, you want to establish behaviours for all of those involved in the process: leaders, vendors, people affected, maybe even customers in some situations. Those who own the project must also own the behaviours. These can then be tracked and become part of the accountability and success metrics.

This goes to the heart of the issue: this is a human process, not a technical one, and that’s where things will either go wrong or right. HBR and industry analyses repeatedly show proof-of-concept “silos” and unclear ownership erode trust and kill scale.

  • Capability (the combined competency to deliver). There are usually four sets of competency involved in change of this kind: The power and potential of the AI itself, which is an entity in this equation, the qualities of the people involved, structured within the strengths of the organisation and its leadership and leveraging the competencies of the vendors providing the solution.

It is the way these components work together that will ultimately determine success. This is real interdependence, and trust is the way that will result in multiplication of capability.  If it feels like it is being done ‘to’ people not ‘with’ them, then there’s a high chance of failure. RAND Corp research shows that AI must augment existing strengths.

In one recent workshop I led, I was told that an organisation now has ‘AI driving licences’ for all of its staff, so that there are agreed and defined ways to use AI and everyone follows the same process. That’s real competency. Empirical research into AI program failures finds capability gaps—skills, governance and integration—are core causes of collapse. Address them deliberately.

What to do differently — four practical moves leaders can make now

  1. Start from your Strategic Narrative, not a vendor demo. Map each AI opportunity back to one of your strategic priorities (purpose, target client, differentiation, value creation – especially the ‘how’). If the mapping is weak, stop or re-scope.

  2. Require an outcomes contract for every pilot. Define what will be better and why this matters: the decision improved, the metric that will change, the experience enhanced, how success will be measured in business terms. McKinsey’s recent surveys show that organisations that treat AI as a value-creation capability (not a string of pilots) get materially better results.

  3. Design for people and consciously identify the right behaviours. Use the Trust Triangle to set behavioural commitments (eg transparency, courage, openness, engagement) so users see AI as part of their work, not an external threat. HBR case studies document how omission of these behaviours stalls adoption.

  4. Focus on the interdependence between those involved as the core of the process. This is a team effort, entering a new world with technology that can go wrong in ways which contains surprises. This requires high levels of trust between those involved so that they can tackle these challenges together and remain focused on the goals in mind (clarity, clarity, clarity). That is helped by governance, collective accountability and well understood role definition. RAND and other studies identify these organisational gaps as the single biggest barrier to moving from proof of concept to scaled impact.

The payoff: sustained advantage, not short term efficiency

When AI is aligned to a clear strategic narrative and introduced through the lens of the Trust Triangle (clarity, character and capability), it ceases to be a plug-in efficiency tool and becomes an amplifier of what makes the organisation special—its chosen clients, its distinct activities, and the people who deliver value. That is the source of sustained competitive advantage. Everything else is just operational delivery – and in that field, any short term gain will be wiped out as others get better too.

Be clear about what makes you special. Leverage AI to make that even more true. And ensure there are high levels of trust between those involved, all clear on the collective ambition, so that you can win together.

Sources

Stuart Maister
Next

When everything constantly changes - what stays the same?