Ian Betteridge Profile Photo

Ian Betteridge

I fought the law and the law won. And then I made up my own law, and it was good.

Business

    ∞

    How to roll out AI in a creative business

    I talked recently about how changing the culture of learning in your business will be important if you will make the most of AI. But no matter what, you’re going to have to roll it out – and you need to do that in a structured way.

    Remember, this isn’t just an ordinary technology roll out: it’s a change management process that will have a lot of impact on your business. One framework which can help, and that I have found incredible powerful for managing change at scale is the ADKAR model of change management.

    This model consists of five stages: Awareness, Desire, Knowledge, Ability, and Reinforcement. Each stage focuses on a different aspect of the change process, from creating a clear vision and generating buy-in, to acquiring the necessary skills and (importantly) sustaining the change over time, something that’s often neglected.

    So how might you use ADKAR when looking at an AI rollout?

    Awareness

    At this point, your focus is to communicate the need and benefits of AI for your business, such as improving efficiency, enhancing customer service, or gaining insights. Explain how AI aligns with your vision, strategy and values, and what challenges it can help you overcome. Use data and evidence to support your case and address any concerns or misconceptions.

    Remember, too, that this stage is about the need for change, not that change is happening. The most important outcome for this stage is that everyone understands the “why”.

    Key elements of building awareness

    • Start with your senior leaders. In effect, you need to go through a managed change process with them first, to ensure they are all aware of the need for change, have a desire to implement it, and have the knowledge they need to do so. Your senior team has probably been through quite a few changes, but none of them will have gone through what you are going to experience with AI.
    • Explain the business drivers making the use of AI essential. Don’t sugar coat this, but be mindful of not using “doom” scenarios. Your model should be Bill Gates’ “Internet Tidal Wave” rather than Stephen Elop’s “Burning Platform”.
    • For every single communication, ask yourself whether it contributes to helping employees be able to think "I understand why this change is needed". If not, rethink that comms.
    • Be clear and consistent in messaging – and have leaders deliver the message (but make sure they are clear about it themselves).
    • Tailor your message. Customize communication for different groups within the organisation. Different stakeholders may have different concerns and questions, so addressing them specifically can be more effective.

    Desire

    Building desire is all about cultivating willingness to support and engage with the change, and for AI, it’s incredibly important. While AI is a technology, it requires cultural change to succeed – and changing a company culture is very hard. Without building desire, any change which threatens the existing culture will fail.

    There are many factors which influence whether you can create a desire for change. Personal circumstances will matter, and the fear with AI is that employees will lose their jobs. That’s a big barrier to building desire.

    And, in some cases, those fears will not be misplaced, so it’s critical to be clear about your plans if you are to win enough trust to create desire. Consider, for example, making a commitment to reskill employees whose roles are affected by AI, rather than giving bland statements about avoiding redundancies “where possible”.

    This is especially critical if you have a poor track record of managing change – so it’s vital that you are in touch with how your change management record really looks to your teams.

    At this point, you should also identify your champions. Who, in the business, has a lot of influence? Who are the people who are at the centre of many things, who act as communicators? Who do other employees go to for help and advice? Are there people who, when a new project starts, are the first names on the list? They are not always senior, so make sure you’re looking across the board for your champions.

    Even if they are not the most senior people or the most engaged with AI at this point, if you win them over and make them part of the project, you will reap the benefits.

    Remember, too, that desire is personal to everyone. While making the business more efficient and profitable tends to get your senior team grinning, not everyone in your business is motivated by that. Focus, too, on the benefits for people’s careers, work/life balance, and especially with AI, freeing up time to do more creative things and less routine work.

    And don’t, whatever you do, talk about how “if we don’t become more efficient, people will lose their jobs”. I’ve seen this approach taken many times, and in creative businesses, it almost never works. Desire is about motivating people to change, and fear is a bad motivator.

    Key elements of building desire for AI:

    • Inspire and engage your team members to participate in the AI adoption process.
    • Identify and involve key influencers and champions who can advocate for AI and influence others.
    • Highlight the personal and professional advantages of AI, such as learning new skills, increasing productivity, or advancing career opportunities.
    • Create a sense of urgency and excitement around AI and its potential.

    Knowledge

    If awareness is about the why, the knowledge stage is about the how: how are we going to use these tools? This is where you build knowledge of the tools and the processes by which you use them.

    One mistake that I have seen made – OK, to be honest, I have made – is to focus too heavily on training people on how to use a tool, without also training on changes in the processes you’re expecting people to make. Every new tool, including AI, comes with processes changes. And, in fact, the process changes that the tool enables are where you achieve the biggest benefits.

    Training people in the context of the processes they follow (and any associated changes) relates the training to what people do – and that’s why I would recommend role-based training, which may cut across teams. If you have large teams, consider further segmenting this according to levels of experience. But I would recommend that you train everyone if possible: people who are left out may end up feeling either that AI isn’t relevant to them (and it will be) or that they have no future in your new, AI-enabled business.

    Key elements of building knowledge of AI:

    • Provide adequate and relevant training and resources for your team members to learn about AI and how to use it effectively. Make sure you document any process changes.
    • Tailor the training to suit different learning styles, levels of expertise, and roles.
    • Use a range of methods, such as workshops, webinars, online courses, or peer coaching.
    • Encourage feedback and evaluation to measure progress and identify gaps.

    Ability

    So far, what we have done is all theory. This stage is where the rubber really hits the road because it’s where all that training starts to be implemented. And at this point, people will start to spot issues they didn't see before as they get the hang of new processes and get better at them. They will also find things you didn’t anticipate, and even better ways of using AI.

    One aspect that’s critical at this stage is the generation of short-term wins. For a lot of your teams, AI is the proverbial big scary thing which is going to cost them their jobs – and even if you have had a successful “desire” phase, it can be easy for people to be knocked off course when that is at the back of their minds, or they are reading scare stories about how AI will mean the end of humanity.

    Quick wins will help with this. They are positive, visible evidence about the success of people they know using AI, and in storytelling terms that is absolute gold dust. Remember, though, that the positives must be personal, and in a creative business they need to focus on improving the creative work. Shaving 10% of the time taken from a boring business process might be incredibly valuable to you, but it’s not all that compelling to a writer, editor, or video producer.  

    Key elements of building ability in AI:

    • Support your team members to apply their AI knowledge and skills in their daily work.
    • Create a safe and supportive environment where they can experiment, practice, and learn from mistakes.
    • Provide guidance, feedback, and recognition to reinforce positive behaviours and outcomes.
    • Make sure success stories are being shared, and that your teams are helping each other.
    • Monitor and track performance and results to ensure quality and consistency.

    Reinforcement

    This stage focuses on activities that help make a change stick and prevent individuals from reverting to old habits or behaviours, and I think it’s both the most crucial stage of managing a change in technology or process – and the one that’s easily forgotten.

    There are several reasons for this. First, commitment even among your senior team may be waning, leading to reduced encouragement from the top to continue along the path. The people who thought that your rollout of AI was likely to fail will probably be latching on to every bump in the road and turning them into roadblocks – ones that they “knew would happen”.

    This is why it’s incredibly important to have all your senior team go through a parallel managed change process, to make sure they are all bought into what you want to achieve. AI is a strategic change on the same level of impact long-term as a complete restructure of your entire business, so there is no getting round managing that process for your senior team.

    If you are starting to get resistance to AI deployment at this stage, check whether your senior team are still bought into it. In the worst case, some of them may be sending subconscious signals to their teams that they don’t have to keep going.

    And now the bad news: in terms of budget, the reinforcement phase may cost as much as the training required in the knowledge phase because you need people looking after the AI roll out who are constantly engaging with your teams, understanding issues, celebrating success, and making sure that communications about how AI works for everyone, and – importantly – keeps everyone updated on new developments and changes.

    For every new pitch, product or process, someone needs to be in the room asking how you can use AI to improve this, speed it up, or do interesting creative things. That is the only way they AI will become embedded in what you do, and not fade away – as so many corporate projects do.

    Who is that person going to be? The likelihood is that in the “desire” phase, internal champions will emerge who can do that job. This offers the advantage of credibility, as it’s someone who is both personally familiar and professionally respected, but don’t make the mistake of assuming this role is something that you can tack on to a day job. Unless your business is very small, doing all this is a full-time role, for at least a year after you have “completed” the rollout of the technology.

    Key elements of reinforcing AI use:

    • Celebrate and reward your team members for their achievements and contributions to the AI adoption process.
    • Focus on improvements in employees’ experience, not just business benefits.
    • Solicit and act on feedback to improve and refine your AI practices and policies.
    • Reinforce the benefits and value of AI for your business and your team.
    • Keep your team informed and updated on the latest AI trends and developments and encourage continuous learning and improvement.
    ∞

    Weeknotes: Sunday 16th August 2020

    I missed last week’s note thanks to a huge bout of tiredness which left me pretty exhausted and sleepy all Sunday. Sorry about that. Still a bit knackered now, so this will be a pretty short one.

    Antitrust is here again

    Back in the mid-noughties I spent a while covering the Microsoft/European Commission antitrust investigate, the one which ultimately led to the “browser” choice” version of Windows (where everyone naturally chose Chrome, because at the time Chrome didn’t suck).

    That meant I had to learn an awful lot of antitrust law, and – as I was writing for an American site – how European rules differ from US ones. The news that Apple is being sued by Epic Games means a whole new generation of technology journalists are about the learn a lot of the same stuff. It’s fun.

    One thing to understand off the bat: in Europe, there’s an assumption that competition is good for consumers, and so things which restrict competition must have a VERY clear consumer benefit. No such assumption exists in the US, where immediate consumer harm is all that really matters.

    This is going to make things pretty tough for Epic, because Apple can ask “where’s the harm?” and Epic needs to do the work to show it. Just a restraint on Epic’s freedom to do what they hell they want won’t be enough. And Apple has a strong case that a single app store with a fixed fee has benefited consumers by providing developers with a clear route to market, as well as something that’s much more secure than mobile app distribution used to be. Anyone who remembers the pre-App Store era will know what a shambles it was trying to get mobile software if you weren’t a nerd.

    Stuff I’ve been reading

    Ars Technica has a great interview with two of Apple’s leading AI experts. It’s worth remember that Apple believes machine learning is so core to what it does that it’s built in specialised ML hardware into its processors for years.

    Meanwhile, Microsoft is all in on cutting its carbon emissions and making itself carbon negative. That’s both aggressive and admirable. Satya Nadella is some leader.

    I’m incredibly proud of my former colleague Thomas McMullan, who has a book coming out. Tom is proper clever and you should read his stuff.