IT adoption is hard at the best of times. Many projects fail not in their implementation, but after in fact – often due to lack of adoption and training to embed the change.
AI potentially poses an even greater challenge as it is new ground for all organisations. With sensationalist headlines around job losses, it’s understandable if your employees view AI with suspicion.
However, to achieve success with your AI initiatives the same principles apply as they always have. It comes back to the culture you create: one that embraces change and where initiatives aren’t enforced top-down but start with the end user in mind.
Don’t ignore the elephant in the room
Almost a third of workers believe their role will be diminished or replaced by AI, according to research by Grant Thorton. In another survey by YouGov, three quarters of people believed AI would lead to job losses in the business world.
Workers are nervous and that can’t be ignored.
One way to allay those fears is to ensure your employees have a clearly defined role aligned with overall business objectives. Fundamentally, a sense of purpose.
Admin and other low-value work aren’t satisfying and will not help the company. Be up front that AI is there to remove that kind of work so they can instead focus on higher priority tasks that are more fulfilling for them and benefit the business.
Why would you spend time in a contact centre doing admin when you could be helping customers understand product options, upselling services or solving issues that may be impacting customer retention? It’s about freeing up time to focus on this more valuable work.
“The customer has a better experience, the employee has a better experience,” said AI thought leader Ant Morse in a recent chat with CloudClevr. “Organisations have a duty to pull the handbrake and say right, let’s just pause. What could it do? What might it do? And what are the benefits on both sides?”
By making roles more productive and beneficial to the business, AI actually makes the role more secure, not less so, and that should be a core message of any AI rollout with your teams.
Start small, get training right, then scale
Starting with a small pilot team will allow you to manage the process more easily and also help ensure success by focusing initial training on a core group of users. Too many, and you risk the change not being embedded properly. Instead, bring the pilot group up to speed, and then they can act as change champions to support a wider rollout.
It’s an approach Access Partnership took with its successful Microsoft Copilot implementation.
That project also highlighted the importance of the right training, even if using an out-of-the-box tool such as Copilot. Access Partnership greatly benefited from guidance they received from an external expert who trained them on prompting. This sort of enablement is important to getting the right results out of AI, whatever the use case.
It’s also worth giving general guidance on AI as a tool – the pros and cons, and how AI is not infallible. This underscores the need for humans to review the output and further illustrates that AI will most often support workers, not replace them.
Knowledge sharing is vital
We recently suggested starting an AI steering group. It’s good to have a core cross-functional team that can focus on AI and share knowledge around successes and failures. This will help your business avoid repeating mistakes and accelerate adoption.
This doesn’t have to be just senior team members; it can be those who use AI tools daily and put them into action as part of their daily roles.
Some companies encourage employees who use AI tools to share their workflows so that best practices can be developed and knowledge can be shared across the organisation. One Reddit user said their company had even created an “AI Council” of power users who shared their knowledge of what was working with other teams.
Access Partnership set up a SharePoint site to document best practices and share their learnings with the wider team, helping users quickly get up to speed.
Finding the right way to do it is part of your AI journey, but sharing that knowledge is vital.
AI might not be right for every process
Not every process will adapt easily to AI, and that’s fine. The job shouldn’t be to apply to AI to every process, it’s to find the processes that will be improved most.
Categorise processes by whether AI will be easy or difficult to implement, and park those where AI isn’t a good fit. Then consider whether AI is a full replacement for the entire process or something to augment employees in their roles.
As a result, you will likely end up creating new processes entirely because the process may be a mixture of AI and human interaction.
Ant Morse refers to it as Cobotics, humans working alongside AI. There could be a situation, for example, that an AI chat bot gets stuck and it needs to pass to a human. So here you’ve just created yourself a new process rather than revising an old one. What is the handover process to a human, what information is shown to them and what happens if an agent isn’t available?
Each of these scenarios needs mapping out so you have an understanding of what you want to achieve in the process. Take fresh eyes to the process and make sure that it’s necessary and optimised. Otherwise, AI will just try to add automation to a bad process.
Don’t go too far too fast
If you try to implement AI to fix too many things too fast, the bigger risk is that it will not achieve your goals, and your employees will be dissatisfied.
Yes, AI may be disruptive as a new technology. However, when it comes to embedding it, no one will welcome a culture of change and disruption.
It’s better to implement it in some key areas, ensure the right training is in place, and tweak it as you go along. Then, when you can demonstrate success, move on to other processes and areas. Not only will you have the learnings to help you with the next stage, you’ll start to create proponents from within the very teams that you are trying to implement AI.
“It’s going to be disruptive, but only if organisations don’t get it right,” said Ant Morse. “We’ve got a window of opportunity here to train people, to evolve, to plan and prepare, which we must take.”
The benefits can be realised, if rolled out with care
AI will be feared by some employees and that won’t change overnight. Acknowledge that, then involve them as the subject matter experts they are.
Change will succeed and embed itself in your organisation if the participants feel they are along for the ride and have some control over the journey. Build AI to solve problems in their lives, not as a top-down initiative, and you will be well placed to reap the rewards of AI.
