Running in fog
Getting your AI strategy right
In the last Moving Tribes post about the challenges of managing change in an organisation, I made reference to the fact that ill-informed management teams in the early days of online retailing were prone to making big but ultimately incorrect investment decisions about digital technology projects that they didn’t really understand.
So its confession time - that’s a club of which I am a member.
I’m not alone, though. When, years later, I told the story at a business dinner about what I’d learned from signing off a multi-million pound investment in a ‘web platform change’ that took far too long and ended up delivering technology that was out of date before it even arrived, I was amazed by the number of people sheepishly raising their hands and confessing to having made the same error.
The root cause of that apparently common mistake was a toxic blend of circumstances. Internet retailing was, at that point, fast growing and obviously very important for the future of any retail business, but it was also new and therefore not particularly well understood by boards or management teams. Even the technology leads in many retail businesses were experts in enterprise systems, tills and other ‘normal’ retail technologies but didn’t necessarily know that much about the web.
The result - when Boards gathered to decide on a digital strategy, the only actual experts in the room were usually the vendors or consultants proposing the project, and that is never a great recipe for a good outcome.
Looking back, this period of ill-advised web investment had clear phases:
“This new technology is not proper retailing, will never catch on and has no place in my business”
“OK, there are clearly a lot of start-ups doing quite well by focussing on that new technology but it is still basically cheating and will never replace my hard-won personal expertise”
“Oh god, we are getting left behind, let’s get a big consulting firm to come up with a plan”
“Let’s try to earn some bragging rights by writing the biggest cheque we can for a huge transformation project to catapult us from laggards to leader”
“Oops”
And finally, after riding that bronco, the world eventually settles down into a series of sensible, better understood and smaller investments which are a lot more likely to actually work.
And now, I suspect, we are about to ride the same rodeo with AI technology. I’ve heard plenty of established retailers explain how it doesn’t really apply to them (phase 1) and seen plenty of over-excited start-ups announce how they will change the world (2).
My worry now is that we are entering the ‘big cheque’ territory of phases 3 and 4. Imagine my horror, then, at the headline “John Lewis bets on AI-powered shopping and TikTok Shop in £800m investment”.
Luckily, this is not quite as bad as it seems when you read the detail, as these AI investments are a part of a much bigger investment programme which I suspect lumps in store refits, wage rises and a good deal else to get the eye-catching headline number.
Nonetheless, there is a danger that we are about to enter the phase where, just like the heady days which you could make your share price go up just by adding “.com” to your name, businesses try to generate investor excitement by being seen to ‘one-up’ each other with their AI announcements.
Perhaps a better alternative for your business might be to learn from the pattern of the past so as not to have to repeat it. A series of small experiments in how AI technologies can work for you, using the ‘test and learn’ philosophy we discussed in the last Moving Tribes, is much more likely to yield results than a single big bang project.
Get good at learning from these kinds of experiments, and then pushing resources quickly behind the ones that seem to deliver results, and you maximise your chances of making the best use of this powerful new tech and avoiding the pitfalls of the hail-Mary approach.
I’ve often likened these periods of rapid change to navigating through a thick fog. You might have a decent map and think you know the direction you should be heading, but however confident you are, it is unwise to simply run off in that direction at full speed. Without being able to see clearly ahead, there is always the danger of running off a cliff or falling down a hole.
Instead, the best approach when the future is uncertain is a careful one, learning from each new landmark that emerges from the fog. That might not get you a round of applause in a conference speech, but I’m fairly sure another “£10m on the wrong project” club is going to emerge from this AI revolution, and you don’t want to join it.