Written by Annette Wilson, Microsoft Implementer & Trainer, Shackleton Technologies
There’s been a lot of conversation recently about the risks and challenges of AI …. and rightly so.
But alongside that, there are also practical, everyday benefits that are sometimes easy to overlook.
From my perspective, it’s not about taking one side or the other. It’s about recognising both, understanding where AI can add real value, and where it needs care, caution, and oversight.
This case study is a good example of that balance in practice. Not about blind adoption or outright resistance, just how use evolves when something genuinely proves useful.
Not everyone jumps straight into AI. In fact, some people actively avoid it – until something shifts.
At Shackleton Technologies, Copilot is already part of how we work day-to-day. Different teams use it in different ways – supporting delivery, analysing information, or taking some of the admin burden away.
There isn’t a single “right” way to use it — and that’s intentional. Adoption tends to happen gradually, and often quite differently depending on the role.
Andrew, one of our Account Managers, is a good example of what that looks like in practice.
When I spoke to Andrew, I’d describe him as slightly sceptical – not dismissive, just unconvinced.
His experience had mostly been with Copilot (or Copilot Chat), and his view was fairly typical:
(Completely fair – and for context, Andrew is more than capable of writing a good email without any help.)
For him, Copilot didn’t feel essential. It was something he could use, rather than something he needed. That said, he didn’t rule it out entirely.
What did resonate was the more practical, low-effort functionality – particularly:
In an account management role – where time is split across clients, meetings, and follow-ups – even small-time savings can make a noticeable difference.
Still, at this stage, it felt more like a helpful extra than something he’d actively rely on.
That changed when Andrew explored Copilot Cowork through the M365 Academy as part of the Frontier programme.
Instead of using AI in short bursts – asking a question here and there in chat – he used it to produce a complete piece of work.
That shift, from interaction to outcome, was key.
The standout feedback was surprisingly simple: “It just does it.”
Behind that comment is something quite important — the difference between assisting with a task and delivering something you can actually work with.
The biggest difference was the nature of what came back.
Instead of fragments or half-finished drafts, Cowork produced:
That changes the experience entirely.
Previously, the process looked like:
With Cowork, it became:
It’s a subtle shift, but a powerful one.
You’re no longer creating everything from scratch – you’re improving something that already exists.
For an Account Manager, this is particularly valuable.
Day-to-day work often includes:
These tasks aren’t difficult – but they are time-consuming.
Starting from a complete draft changes the dynamic.
Instead of building, the focus moves to:
That reduces friction and frees up time for higher-value work – particularly client-facing activity.
Another noticeable shift was in the quality of the outputs. Andrew described the documents as:
In some cases, they also included supporting visuals, which made them feel more like finished pieces of work rather than early drafts.
It wasn’t perfect – and he wouldn’t claim it is. But it was a significant step up from the disjointed, sometimes frustrating experience he’d had previously with chat-based interactions.
One aspect that stood out was visibility.
Being able to review:
…added a level of transparency.
It didn’t feel like a black box producing an answer. There was a clearer sense of how the output had been created. For someone who had previously questioned the reliability of AI outputs, that visibility made a real difference. It helped build confidence in what he was reviewing and, importantly, what he was choosing to use.
A few things really stood out as making the difference:
What’s interesting is that this didn’t completely transform Andrew overnight. He hasn’t suddenly become an AI enthusiast. There’s still a level of caution and a healthy amount of questioning. And honestly, that’s exactly what you’d want. Adoption doesn’t have to be all-or-nothing.
What has changed is how he uses it. He’s moved from limited, occasional use to using it deliberately, where it genuinely adds value.
Across the business, that’s where Copilot works best:
In Andrew’s case, that meant moving beyond chat and into something more complete and outcome focused.
And it highlights where Copilot Cowork really adds value:
Moving from prompts and ideas To complete, usable outputs
As with any new capability, there’s also a practical consideration around how it’s used – particularly as tools like Copilot Cowork move towards usage-based models [1] [2].
In practice, that tends to shift the focus slightly towards where it genuinely adds value, and how it fits into day-to-day work.
Less about whether to use it or not, and more about using it deliberately, in the right places.
#M365Academy #ArtificialIntelligence #AIAdoption #Microsoft365 #CopilotCowork #FutureOfWork #ResponsibleAI #ShackletonTechnologies
Sources: [1] Microsoft.com/Cowork [2] Windows Forum
Click below and ‘share’ this article!