Welcome to the AI Early Adopter Club™

However much I use generative AI day-to-day, I still have moments when it feels like pure wizardry. Those are the moments that remind me we are just getting started on the big innovation adoption curve that you still see featured in workshop slide decks the world over. And that is how we find ourselves revisiting 1962.

That rings a bell

The Diffusion of Innovation theory, probably more familiar as the concept of a market adoption bell curve or change curve, explains how new ideas, technologies, or products spread through a population over time. It’s memorable because it’s simple and makes sense. It’s useful because it helps you plan marketing and distribution or – if you’re leading an organisation that is trying to adopt new ways of working – it tells you a lot about how you run effective change programmes.

The generative AI revolution isn’t our first rodeo with disruptive tech, which is why this still feels right. Rogers wrote a lot about farming equipment but his contemporaries – presumably for the same reason that I’m about to skip over the agricultural case studies (my total subject matter ignorance) – often explained it in terms of adopting colour TV (now we’re talking). Whatever your particular point of reference, the model makes it easier to start to think about different groups of adopters, imagine what they’re saying, understand their motivations, and figure out how you get them on board.

The reason Rogers’ model made the leap from, I’m guessing, an actual slide in the slide carousel of a 1960s lecture theatre projector, to everyone’s Change 101 PowerPoint deck is that it is still a powerful way to understand how new ideas and technologies spread.

To re-cap – or explain, if you haven’t come across it before – the model categorises adopters into five groups, distributed across a bell curve: innovators, early adopters, early majority, late majority, and laggards. Understanding these groups is going to be the key to getting your shiny new AI tools embraced by any organisation.

The innovators (2.5%)

The fearless explorers of the tech world, constantly chasing the next big thing… even if it’s still a bit rubbish and they’re not sure what it’s for. These folks thrive on experimentation and need minimal – if any – handholding. They love being first and that’s usually enough to motivate them. They don’t mind a few hiccups along the way.

They can be easy to poke fun at when they are queueing outside the Apple store, as if Apple aren’t going to make enough phones for everyone who wants to pay for one, but we need these people. They will find the bugs and fix them so the rest of us don’t have to. Thank them.

Our top tip: Give these pioneers powerful tools, beta features, and opportunities to experiment. Don’t make them wait or they’ll find somewhere else to play. They’ll figure things out on their own and help you refine your offer. But don’t let them absorb all your time just because they are the keenest to engage: make space for them to discover the next thing, and for you to move on to…

The early adopters (13.5%)

This savvy gang pride themselves on spotting the next big thing before it goes mainstream, and they want to lead change. I know this is you, because gen AI adoption in work is still below 20% and you’re reading this. But you need to find others like you in your teams because they will be your champions – and there’s enough of us that our team of champions can be diverse and therefore influential as you scale up. Engage your early adopters in pilots and they’ll spread the word like wildfire.

These are probably the kind of people who have already bought electric cars because, even though they have a few worries about the infrastructure, they accept the case that science has made for the environment, and they know the battery life is better than it was. These are the people using gen AI for work right now.

Our top tip: Involve early adopters in pilots and provide glowing success stories. They'll be your loudest advocates. Invest time in them and let them invest time in their own learning and development. Ask them if it’s time to stop or time to move onwards and upwards to…

The early majority (34%)

Pragmatic and sensible: they like their tech battle-tested before they commit. Show them clear evidence that generative AI saves time, boosts productivity, or improves results. Give them data for their minds and stories for their hearts. Give them tools that help them in their day job. Keep them on board by removing obstacles promptly.

The early majority are the people who embraced colour TVs once prices dropped and broadcasts improved, which is more than fair, but that’s them in a nutshell. They started buying electric cars only when charging points became more common. They’re in most of the lampposts on my street now. The charging points, not the early majority.

Our top tip: Provide clear, practical examples of how generative AI makes life easier. Solid case studies and peer endorsements will work wonders. Their feedback will be high value and less prone to optimism bias, so pay attention to it. Next up:

The late majority (34%)

They’ll jump on board when their neighbour’s grandma starts using it and swears it’s easy. They’ll need reassurance, more practice, and a serving of peer pressure.

The late majority are the folks who finally switched to colour TVs when their black and white telly broke and they found it hard to replace like-for-like. They’re driving petrol cars right now – or maybe a second-hand Prius.

Our top tip: Realistically, this group needs more hand-holding when it comes to new things. Demonstrate safety, reliability, and ease of use. Seeing their peers succeed is key. Shock and awe will not work; incentives will. They like value. They’ll help you prepare for…

The laggards (16%)

Ah, yes. The ‘my fax machine works just fine’ crowd. Change generally arrives when they’re out of options, or by accident. Removing alternatives is often the only way to get them to switch.

In the Bronze Age, they clung to copper tools long after bronze became the norm, likely muttering, "Copper was good enough for my father..." A few generations later, they railed against electric lighting.

They kept a black and white telly for upstairs much longer than you think, accidentally cheering the wrong team when they scored a goal. It’s probably still in the loft.

Our top tip: Don’t waste time and energy arguing with this group too early on. Make space for the pitch that wins over the majority first, just have in mind that you need to manage the risks and the noise. Sometimes, the only way to bring around a laggard is to remove their familiar alternatives; peer pressure might only make them dig in. But no one really wants to be the end of the bell curve. Or the bell-end, if you will.

Variations on the theme

You know a model is going to last when – like a perfect 90s riff – it’s sampled to create a new classic for modern times. It’s quite common to see the market share s-curve – another 60s classic - applied alongside it. And that works too.

It was as late as 1991 when Geoffrey Moore introduced the concept of a chasm between the early adopters and the early majority in his book “Crossing the Chasm.”

Moore’s theory was about the challenge of making the leap to mainstream (whereas Rogers believed innovativeness was more of a continuous variable). I think they’re both right. Sometimes you see a bit of tech leap from one side of the chasm to the other side overnight after teetering on the brink for a long, long time (I offer you Zoom or Teams during the Covid pandemic). Other times, change creeps up on everyone (like the use of hybrid seed corn by farmers in Iowa, according to Rogers, or – more familiarly – the use of smartphones).

Knowing the pattern is one thing, using it is another

Rogers’ theory - or at least the AI summary I generated and the discussion I partially remember from a Leading Change workshop in a windowless basement in the early 2000s - is about how understanding the market helps you with roll-out. What you really need to know is that his theory also outlines a five-stage decision-making process for everyone adopting innovation:

1. Knowledge/awareness: Becoming aware of the innovation.

2. Persuasion: Forming a favourable or unfavourable attitude toward the innovation.

3. Decision: Weighing the pros and cons and deciding whether to adopt.

4. Implementation: Putting the innovation into use.

5. Confirmation: Evaluating the results of the innovation and deciding whether to continue using it.

Also – crucially – there are influencing factors: The diffusion process is influenced by several things, including the innovation itself, communication channels, time, and the social system. Or, more rarely, a pandemic. But this is in danger of becoming an essay so you can do further reading on your own time. Then ask an AI to act as an expert in innovation diffusion and get it to look over your marketing and change management plans. Ideally, with a private GPT that looks after your data and your IP.

To re-cap

1. Don't sweat the laggards: Start with innovators and early adopters to build momentum.

2. Early adopters = free marketing: Give them success stories to tell.

3. The late majority love receipts: Show them solid proof that generative AI works.

4. For laggards, resistance is futile: If all else fails, make change inevitable.

‍