Six things I believe about Ai and how businesses adopt it
None of this is theory. It is what I have seen work, and fail, inside real UK businesses over the last few years.
Governance before gadgets
Ai without governance is an expensive toy. Before you choose a tool, you need to be able to answer three plain questions: who owns this, what are we measuring, and what do we stop doing if it is not working?
This is where information security and Ai management sit together. ISO 27001 controls the data your Ai runs on. ISO 42001 governs how the Ai operates on it. Skip the first and your Ai governance is built on sand. Get both right and you can move quickly, because everyone knows what is safe to use and what needs a human in the loop.
Read more on thisProcess before technology
You cannot automate a process you have not designed. Most businesses skip this step because they are busy fighting today’s competitors and there is no time to do the unglamorous work of mapping how things actually flow.
So Ai gets bolted onto the mess. One person uses one tool, someone else uses another, two people produce the same report and get different answers because there is no agreed process underneath. That is not an Ai failure. It is a process failure that Ai has amplified. Map the process, fix what is broken, then choose the tool.
Read more on thisOperating systems, not one-off projects
The thing I find missing in most businesses is not capability or ambition. It is systemisation: a coherent way of bringing ISO principles, Lean Six Sigma, project management, change management and technology together in the right order.
Individually these are well known. In combination they are rare, and the businesses that combine them quietly outperform the ones that pick one or two and treat the rest as optional. Joined-up systems beat scattered tools every time, which is exactly why a pile of disconnected Ai apps rarely adds up to a transformation.
Focus is a superpower
A target is not a strategy. When a leadership team tells me their strategy is to increase profit by ten percent, that is a goal wearing a strategy’s clothes. A real strategy is a clear diagnosis, a guiding policy, and a coherent set of actions, in Richard Rumelt’s framing.
The discipline is deciding what you are not going to do. Most Ai pilots fail not because the technology is wrong but because the strategy underneath them is missing. Kill the vanity projects, pick the few moves that actually matter, and put them somewhere the whole team can see.
Human-centred, ethical Ai
The mundane back-office work that soaks up most of a person’s day, the data entry, the document handling, the repetitive queries, is exactly what Ai is best at taking on. Use it well and you free your people up to be the thing they cannot be replaced as: the human connection your customer cannot get anywhere else.
Look after your staff, they look after your customers, and your customers look after your profits, as Richard Branson puts it. I hold to the triple bottom line, profit, planet, people, in that order. You need the profit to sustain the business, and from there the other two are not optional luxuries.
Make yourself redundant
I was trained to make myself redundant, and I apply that at the level of a whole industry. Through teaching and apprenticeships I am helping to build the next generation of internal Ai champions, the people inside a business who will eventually do this work without needing an outside consultant.
I am comfortable with that. The fewer businesses that need someone like me, the further the whole ecosystem moves forward, and the sooner I get to move on to the next problem worth solving.
How ready is your business for Ai?
Get your free readiness score in ten minutes.