I was at a company recently and they were baffled about something. They had a brilliant AI tool that made one area of their business dramatically more efficient. One office was wildly positive about it and the other office simply said, “This doesn’t work for the way we do things.” For this reason, even though the tool had been delivering great results for two years, they still hadn’t rolled it out to anyone on a permanent basis.
This gets to the fundamental problem with AI programmes inside companies. The problem isn’t the tech. The real problem is adoption and, in most cases, the project failed before it even started because they didn’t have one simple thing: an actual AI strategy.
Why an AI Strategy is fundamental
The mistake starts because people don’t understand the impact AI is going to have on their business because they simply aren’t willing to ask the really hard questions that challenge the existence of their current business model. This is not a new problem. It’s what Clayton Christensen talks about in The Innovator’s Dilemma, but AI is exposing it anew as companies face the reality of transforming their business, ways of working, operating model, how they write code, pricing model and perhaps even what they do.
Without this level of thinking, you will typically use AI to make a small number of things a little bit better. Your team will be able to get through 10% more work, and you’ll see some positive results, but you won’t have transformed your business. Using an old example to transform eggs into omelettes, you can’t just leave the eggs next to a hot pan and hope for the best. For many companies, this seems to be the approach to AI transformation: give people Claude and hope the business transforms itself.
Cut, Grow or reinvent
Within the strategy, the company has to decide what the real outcome of the transformation is going to be. One of the most common, but also the least inspirational, is: ‘we can do the same amount of work with fewer people for less money’. That is fantastically dull and extremely short-sighted. The problem is that, as the technology matures, there will be no defensible moat. It will become easier to automate basic tasks, so a company with a low-cost base for the same work will face a race to the bottom where providers are consistently chasing a lower price for the same service, often until the margin completely evaporates. This has happened in almost every low-cost initiative in the past and will happen to any company where low cost is the primary motivator.
The second-order idea is that the company uses AI to focus more resources on growth while staying fundamentally the same. This is really the first strategy in disguise. The basic hypothesis is that you will do the same work for less, giving you more money to invest in growth. It suffers from the same problem as the first strategy and will relatively quickly top out as a viable outcome.
The best strategies force the company to take a long, hard look in the mirror and work out how they are going to reinvent themselves. If they’re an accountant, how might they shift from being a glorified bookkeeper to a trusted business partner? For a software company, how do they solve more problems for their customers and solve them more completely? How will your business model change if 95% of the users of the software are AI agents in the future?
These are hard questions and they can only be answered by having deep knowledge of the technology and your sector, while also being brave. If you haven’t done the hard work of understanding AI, then it’s going to be difficult to arrive at a good solution.
The CEO leads, a dedicated role delivers
Often, a CEO believes that because this is technology, the person responsible for the AI strategy is the CTO. They ask them to produce something and then the CTO does a good (or bad) job before being tasked with delivering the programme of work. Things get built and… nothing changes. The company’s immune system rejects the new tools and ways of working, and nothing gets adopted. This is caused by two significant problems: the CEO mandate and the transformation lead.
If the AI strategy is created by the CTO, people won’t believe it applies to them and it will typically be understood as something done ‘by tech for tech’. If the AI Strategy is done properly and requires the whole business to transform its business model, then it requires the CEO to coordinate all the elements of the business to make that a reality. An AI strategy that is not owned by the CEO cannot succeed in making the business significantly better.
Secondly, the CTO has no authority to get other teams to do what they want. They also typically have a raft of other responsibilities and expecting them to be a transformation specialist is wildly unreasonable. The act of implementing change after a tool has been created is the hard work. It involves training people, listening to complaints and using a whole range of techniques to help people change.
Crawl, Walk and then Run
Having said all of this, I think it’s important to understand the three main tiers of AI transformation and I strongly believe that you need to go through them in order (at least to some degree) if you want to achieve maximum results. It’s always important to remember that you have to walk before you can run.
Tier 1 is simply rolling out the model. That is, you roll out Claude, ChatGPT or Gemini. To be clear, you don’t just hand over the tool. You need to connect your data sources and implement training. For some champions and senior people, genuine face-to-face training with an expert should be provided. For the wider team, training should be a gate. You provide the training, and the team shouldn’t get access to the tool until they’ve passed some kind of quiz or test to make sure they’ve absorbed it. You then need to monitor usage and provide ongoing training to help your team get the best from the products.
Tier 2 is where you implement common AI tools. The number of these tools varies greatly depending on your business, but there are several pre-built tools that solve problems many companies have. Implementing these tools should be quick, simple and add more value than the investment you put in.
Tier 3 is where things get interesting and where the biggest value often lies. This is where you build agents, tools and apps that solve your problems in the way you want your business to solve them. If you do this right, you will be able to provide your product or service in a truly unique way and build an advantage that genuinely compounds over time. It also means that companies which have never built software before may now need to hire a small number of software engineers to build the tools and agents they need.
A Roadmap to Success
Once you’ve determined the strategy, secured the backing of the CEO and understood the tiers, it’s critical that you create a roadmap to share with the business, showing where you are going and in what order. Without this, you’ll get railroaded into the wrong thing at the wrong time. People won’t understand that it’s simply not the right time yet, and it’s common for everyone to lose faith in the direction you’re setting. A roadmap does not need to be complicated or committed to exact delivery timelines, but it does need to establish the order and reasoning behind what you are doing.
Operational Transformation as a core competency
Traditionally, a transformation is a discrete task that ends once a goal has been reached. With AI, I firmly believe that transformation now needs to become a genuine core competency because it will never be done. We are only scratching the surface of what AI will be able to do, and the early versions of what you build will not solve your problems forever. As every other company and the market continue to shift faster with AI, you’ll find that what’s expected of you changes faster than ever before. That means that the second your transformation is ‘done’, the next wave of change will come over the hill and you’ll need to drive the transformation to the next level.
This is why most companies will need to make AI-based operational transformation a genuine core competency that drives the company forward over time.
