How Do You Get Employees to Use Microsoft Copilot?

Alex: Okay, Hannah, thank you. Do you want to start off just telling us who you are and a little bit about yourself?

Hannah: Yeah, absolutely. I'm Hannah Adams. I am head of business learning or learning and development and most recently I was working at ASOS for the last eight years.

Alex: Great. And you're talking to us today around, the rollout of AI and in particular Copilot across ASOS.

Hannah: That's right.

Alex: So do you want to just talk us a little bit about what your role was, within that project?

Hannah: Yeah, absolutely. And interestingly, I had quite a few roles in the project. I first of all started off as a customer, to the first, Copilot 365 license rollout, around.

A hundred, I think of us. Were given access maybe a few more. And we were shown how to use Copilot 365 in the different, Microsoft programs. It was really new to a lot of us. Um. And, really quite different at the time. I'd say it was probably like 18 months to two years ago. So then kind of, I can tell you a bit more about that in a minute.

Then my role changed into thinking, well, how can I kind of make this a bit more accessible for other people? How can I make it more applicable? And a little less scary? 'cause remember, think back 18 months ago, that was when everyone was still saying, AI is gonna take my job.

Alex: And was the rollout initially coming out of it? Was, where was the project

Hannah: Yeah. Yeah. So, the rollout came from the tech team. They've got a fantastic team still, doing kind of AI rollouts and, the initial theory was that if the leaders, got the. AI kind of learning first that they could adopt it and then trickle it down to, their reports. The only hiccup to this being that the leaders are pretty busy and then thinking about how they can also kind of cascade this information down.

It was successful for some areas for sure, but for some areas I think. It was tricky to think about, well, where would I apply that in my day to day? And possibly the leaders didn't have the time for experimentation or the curiosity for experimentation that other user groups might have.

Alex: And so tell us a little bit more about your background and your role, as the head of learning and the skills that that enabled you to bring to the

Hannah: Yeah. Well, I think. So my background is initially not in tech at all. It's in buying. So I'm a retail person through and through, and having been at ASOS for eight years and working on kind of lots of transformation projects as part of the business learning and new starter learning and then different learning interventions.

Um. I knew intrinsically that you have to kind of adapt learning to different, functions and different roles to get it to stick so. Anything that you can do to understand what people do in their role, what they spend a lot of time on, where their pinch points are. What's really repetitive and time consuming is when if you introduce something that addresses those problems.

That's when you're gonna get the most success. So I kind of applied some of that thinking to this, you know, we've got this great new tool, got this really exciting, Copilot suite. How are we gonna get, more people using it, not even the 365, just kind of like Copilot web and teams premium.

Like how can we even let people know that they can use that within their roles?

Alex: So you touched on earlier, you had a little bit of a false start as, as an organisation in terms of that rollout. So just dive into that a little bit more around what the, what the original approach was and really where it, where it failed.

Hannah: Yeah, I think. There were some successes, don't get me wrong. Some people walked away from that learning and they were really excited and started using Copilot for work and leaving where we'd all started just using ChatGPT for probably personal use. They started using it for work. So there were definitely a few of us.

Um. But the approach really improved when the project team had a rethink, and the rethink was, let's reach out to people who are really enthusiastic about. AI and who like talking to their communities about it. So the team who were kind of completing the rollout created, a really great collective of, people from different functions across the business and that really helped to embed.

That kind of community spirit once again, you know, it's all about the people, and not just the kind of the tool itself. People thinking about different use cases where they could use it in the day to day, and that's what really inspired me to think. Right. I. Okay, what use cases can I think of? How can I help my team to adapt this?

How can I help new starters, to pick this up and give them best practice when they join? And how can I help like, really busy people in the organisation with some of their kind of lower value, work to kind of. Spend more time on higher value work. So I kind of started targeting, and having conversations with different, functions and different user groups, which really seemed to help.

Alex: And culturally, where would you say a os were when they went into this? Skeptical of AI or were they keen to embrace it, but just a bit kind of unsure how to use

Hannah: Definitely the latter. So, myself and my team had run a learning needs analysis, annually, and we'd identified that our, PA and EA population, have to work hard and fast, have to support a lot of the leadership and the exec, and there were some really great efficiencies that they could gain from using Copilot in.

In their work, so. Using, an SME and sponsor within the EA population, who's just fantastic. We designed a learning program of showing them different AI tools based on examples that they have, like using Teams premium, for example, to record a meeting so it takes notes for you. Or, being able to kind of catch up on emails within Outlook, once they got stacked up for the day or.

This kind of goes outside Copilot a little bit, but kind of using Miro to support on workshop and whiteboarding sessions and using the AI and that to kind of group things together. So once, myself and the two brilliant guys, who worked on this with me showed them how easy it was and got them in to have a go because I think that's another big part, don't you?

Of like, just have a play, you're not gonna break it. Got them in the system to have a go. They were like, oh, okay. This is really gonna help me. It's gonna help my leader and this is gonna free me up to do other cool stuff. I think PowerPoint was a really interesting one where they'll spend a lot of time making PowerPoints for people and and it's really time consuming.

And by using the kind of AI within PowerPoint, it really kind of sped up, sped up what they're able to do.

Alex: So they effectively became your advocates.

Hannah: Yes.

Alex: Of

Hannah: Yes, yes. So the really great thing, by the way, about the PA and EA advocates was that we were able, they kind of spoke to their leaders about it. So whereas the initial approach was to speak to the leaders first by then having a. The leaders, support team say to them, this is so great. Have you thought about using this?

I've done this. It's so brilliant. It really helped to influence from a different message. So, the next group was new starters, myself and, a great guy I worked with. Raz built a intro to, AI for new starters. So any new starter ASOS would get a. Copilot web and teams premium learning in their kind of first or second week, and that meant that even if their team weren't using it at the time, that the new starter would be able to come in and know the best practice for calls and for safely and securely.

Prompting within, Copilot web, and that went down really well. We even ended up running that session across, you know, inviting people who are existing in HR to come along. And then I asked a couple of people from within the buying and merchandising area to check it out, and they were like, wow.

We didn't know. We didn't know this was possible. And buying a merchandising is quite a big area. So then we started going to some of their meetings to say, these are some of the things you can do. And then the kind of the different use cases escalated, which I can, I can tell you about a bit more in a minute.

Alex: Oh, that's great. I love the idea of having the new starters come in and do

Hannah: Yes,

Alex: It makes real sense because they've probably got a bit more time, when they're, when they're joining to learn it, but also they haven't got old ways of working that need changing. How, how did that sit with the existing team when the new starters were coming in and benefiting from all the things that AI bring?

Hannah: I guess it went well 'cause I never heard a single complaint. And. Because other people in the business started asking like, what's this course you've been running? And, have we got this in our, in our pc? I was like, yeah, yeah. It kind of got everyone quite excited, I think, just to understand, what was at their fingertips.

Um. So the next step, which was really cool within kind of design and buying a merchandising, was just to show them how they could go further than writing kind of the simple prompts of getting started with stuff. So in buying a merchandising, you have to write a lot of strategies. You have to, kind of present quite a lot.

So we said, well. Try asking it how to prepare for a strategy, but explain in your, and we had to obviously explain what a prompt was, explain what the purpose of the meeting is, what you are hoping to get out of it, how you want to come across. And then we got them once again to have a go and they're like, oh my God, this is really good.

And they're like, oh, but that bit's wrong. That bit sounds really odd. So we were like, okay, so explain what you wanna be changed. So then they would go and re-prompt and they were like, oh, that's better. So I think that then opened up a lot of. A lot of inspiration and the design team, we showed them how you could look at different trends that were on the catwalk and you could say, Hey, tell me from the Paris, September.

Catwalks for women's wear. What were the key colors and trends and fabrics? Who were the designers that got talked about a lot? So then you could kind of prompt again, like show me some images. And these things would usually take the teams a really long time to troll through, and I'm sure they're still doing that really thorough research.

But if you want a kind of quick snapshot. In your meeting that day, you can say, this is what Paris has been all about this time. You've, you can prompt it quite quickly. I guess it just goes back to that caveat, doesn't it, Alex, of checking that the information you are getting is on point and you are not kind of giving, trend information that that isn't accurate.

But, after the tool gets to know you well. You can obviously kind of prompt it with a, with a bit more trust and confidence.

Alex: And you gave some really nice examples there, and I think, it was the handholding part of that that's key because think we've all seen that reaction that you spoke about of, well that looks good, but this bit's wrong. Or like you say that you might experience a bit of hallucination. I think that's the point that a lot of people give up and go, oh, well actually it's not very good. Um. But actually you're able to kind of work through that and say, okay, why isn't that? Let's work with the tool to rectify that so it can learn, and so that you can learn how to use it better and also understand again, that it's, it's a supporting role in you doing your job. You're not outsourcing it

Hannah: Yeah, completely.

Alex: Do a hundred percent. How did you, how did you go about approaching that and how do you think others could sort of learn from that approach?

Hannah: I ended up having a bit of a mantra, which was, can AI help me with that? So getting people just to think about when they're turning on their computer every day. Rather than going through every email one by one, can you pull up a summary, rather than using your trustee notebook to take notes, can you just turn on record and get your AI notes from it?

One of my team, they almost challenge themselves to find thing, different cases. So. They were like, oh, we need to find a bit of code to go in the back of this, Power BI that we are building and we dunno how to do it. And they said, we are gonna, we are gonna ask Copilot and see if we can get it.

So they actually really challenged themselves to almost get a bit of code that they, 'cause they're not engineers, they couldn't create themselves, but they managed to get it. And then when they kind of plugged it in and they got the, Power BI working how they wanted to, they were so pleased. So. It's kind of encouraging everyone to like, go above and beyond the simple use cases that you've given for your area and just kind of think, oh, well, you know, let's give it a go.

Let's, let's try writing a decent prompt and see what I can get back.

Alex: And, and how far did that terms of. The usage. So obviously if you're starting with experimenting and

Hannah: Hmm.

Alex: Prompting, were people also starting to look at how they could build out agents and how they could kind of use it, more substantially going forward.

Hannah: Yeah, definitely. So, I wasn't involved in this one, but I know that, uh. There were a couple of teams in customer care and finance who were looking at kind of a lot more automation and agents. And then, one of the fab guys on the project helped us to build a bot for new starters. So new starters again, often have a lot of similar queries.

So, they would usually go to, a shared inbox. And people would have to get back to them. But by creating the new starter bot with, feeds of different kind of common queries and common answers, when the bot was implemented, it really reduced those kinds of queries, leaving the people who are kind of looking after the shared inbox with the new starts queries to do those more complex.

Complicated kind of queries that a new starter might have. Rather than, how do I check my holiday? Or where can I find my new starter? Learning? They were able to kind of deal with those bigger ticket items.

Alex: Great. And thinking around the comms of the project. What were the messages, and the communications coming, I guess, outta the business in, into, employees, and was driving that? Who was making that communication?

Hannah: Yeah, I think the original comms that the team did around. They had this work Smarter campaign. It was really enticing and interesting and it showed lots of what the different tools could do, including like a really cool animated character, which depicted AI and I dunno if you've ever seen in the ASOS head office before, but we've got lots of big screens in there.

So it was on the big screens and, there was a whole, internal comms site that was. Created with like high level detail, but then like bigger amounts of detail that comms would come off there. And like I said earlier, we had that community, as well within teams where we used to kind of share and ideate as well.

So, I'd say that's where a lot of it came from while I was still there. But I think what really started to take off was, once again, when we went locally. If that's been a teams meeting, then you have people saying on the chat, oh, how did you do that thing again? And then before I even answered, one of the others would would say, oh, this is how you did it.

Oh, and by the way, I tried this. Have you tried that? So I think. Company wide comms are cool, but when you start going into functions and roles with similar people who do similar jobs, and who are on a similar kind of stage of the journey, that's when you get that really meaningful dialogue of just kind of.

I guess embedding that people are just kind of bring it into their every day and really building their confidence with it. So I think teams chats, slack chats, just getting that kind of community is definitely a way to, take it from company-wide to function or kind of business unit.

Alex: And I'm, I'm gonna ask you to tell tales now. Who, who were the, who were the laggards? Who was, who was the troublemakers that wouldn't, wouldn't quite go along with it all.

Hannah: Mm, I guess.

I guess it was the people who didn't have the time to experiment. Who had, or who had a set way of doing something that they were really comfortable with. And if you've been in your role for 20 years and you've always done something in the same way, you're kinda like, oh, well, why change it now? So, or people that kind of liked using.

Chat, GPT. And they liked it for the fact you could speak into it and you could kind of use it for off on the hoof stuff. But they were less, they found it less easy to adapt to. Well, you can kind of put more into there or you can kind of use it in your Excel or you can use it in your power bi. So. I dunno if it was a particular function or maybe, as we know with the change curve, there's always some keynote easy, like early adopters.

And then there's some other people who really need persuading and they almost need lots of other people around them to be using it to the point where they think, okay, I have to do it now, don't I?

Alex: Yeah, and I can sympathize with those and I think, like you say, the time poor is kind of one of those key things. 'cause. The first time you're doing things often, it does take longer using AI because you have the back and forth or you want to build out a custom GPT and you think, well, it would be quicker to just do it. But it's that, longer term benefit of efficiency that you gain and you kind of almost have to work through that barrier. But that is hard when you, when you've got a day job.

Hannah: Yeah, and I think especially if you are working in a really heavily tech company, but you're not a tech person yourself, so. A tech person would kind of have known about all of this stuff for a lot longer and feel more comfortable with it. But if you are in a role where producing clothes, or that kind of thing is more what you are used to, it's gonna take a bit longer for you to trust a new tool.

Alex: Yeah. Yeah, naturally. Um,

Hannah: I.

Alex: What do you think have been the biggest lessons from the rollout?

Hannah: I think just getting really human talking to people, kind of chatting to people and kind of showing them what it can do. But honestly, once a few people kind of knew what it could do and then others got wind of it, then. It caught on quite quickly in certain areas, and now I'm pretty sure there aren't many areas of the business that aren't using it.

So I think.

I just think getting people in there and building the confidence and spending, like you said earlier on, that kind of actual like learning lab, have a go with people is what's, is what's really key.

Alex: And if someone's listening and they're about to embark on a rollout across their organisation, what would be the the practical advice that you'd give them?

Hannah: Do a bit of a learning needs analysis. Find out what people do in their roles. So I guess legal work with big contracts all the time and we'll need to compare them a lot and that they'll probably work in Word and that kind of thing a lot, so that will be more prevalent for them. But in finance, it'll be a different kettle of fish with all of the data they use.

And once again, like HR will be working with lots of confidential documentation. So they'll probably have nervousness around that. But once again, lots of, lots of data and lots of written words. So I guess it's sitting down with your stakeholders and your subject matter experts in function to find out.

Where are you using AI at the moment? What's stopping you from using it? Like describe to me like high areas of work and what you spend a lot of time on. Asking them what they'd be open to trying with, because people have got to be comfortable. Right. And then like we said earlier, once you get a win, then people are gonna try a bit more.

So definitely. Not only speaking to leadership, but seeing if you can get a champion in your, in your function. Like we had our digital collective, that used to champion and they would pipe up in team meetings and say, I've learned this great new thing, or put a link into the team chat, someone who's vocal and really curious, will be really great for your network.

And the new starters, I'd say kind of catch them early. While they're, while they're fresh to the business, because then they won't be relying on someone else to teach them. They will kind of know what's available and they can just kind of teach themselves as they go.

Alex: And part of what you touched on there was around making it. And specific to

Hannah: Mm-hmm.

Alex: And to their roles. A lot of people listening probably wouldn't have ever done, a learning needs analysis. I certainly don't think I have, or if I have, I've never called it that. Have you got some tips or a framework of how a business, or sorry, how an individual could approach doing.

Hannah: I would probably prompt it. And my prompt would be, I am a learning professional. I want to run a learning needs analysis for my organisation to, help kind of understand what their day-to-day roles and bottlenecks are. Can you give me a great format? Um. And kind of order of activities that I need to run.

That's probably where I would start because, and it will say something like, identify your stakeholders. Understand. Like set up meetings to understand what they're doing, interview them and take notes. Capture your key findings, and then based on your key findings and I guess where you are in the project of how quickly you need to roll it out and to whom it'll give you the priorities of your most either low hanging fruit areas or areas that are the biggest, that need it the most.

And then you and your project team can kind of look at, well. Where do we think based on our business strategy and the needs, where should we get started or can we maybe do like one big area and one smaller area and run them at the same time, and then compare notes afterwards? That's the other thing as well, Alex.

I'd probably never try and roll it out to everyone at once. I would kind of do a couple of teams or a couple of functions at a time. Because then you get your learns as you go along and you can be quite agile and you can iterate your approach. So what doesn't work as well for one group or you know that you could have kind of done something more effectively.

You can definitely apply to the next group. I.

Alex: And you sort of talked to yourself out of a job there and just said, AI can, can do it if you prompt it. But obviously we know that's not the reality. And, and you have the depth of the subject matter expert. How are you using AI so that it doesn't just replace you? I.

Hannah: I mean from a learning point of view at the moment, I'd say I love it for starting on a blank page like so many people do. If I am looking at a new subject that I'm working on, I will write a really detailed prompt about what. What it is, what I'm hoping to achieve, what I already know and ask for kind of an idea of a framework and some cool activities, but that's kind of where.

The AI's and I begin after a really hefty prompt. I think it's to then kind of personalize it to the business you're working with. Personalize it to the industry, to the people in it. And it's with that bespoke personalization that you're gonna get more success because what works for finance doesn't work for legal.

And what works for design isn't gonna work for hr. So. It's kind of having that framework and then understanding with your human skills, the flexibility that you need, and you can get all of that from your results of your learning needs analysis to then kind of flex your program as you go. I.

Alex: And where do you see the opportunities for using AI going forward in in your career?

Hannah: I mean, you and I were a absolutely brilliant, event last week. Was it? With, the AI change makers and I was definitely thinking while I was there about different tools to create videos, to create exercises, checklists, more interaction with learning. 'cause we know that. Although people really love face-to-face learning, it can be hard getting everyone together in the same place.

When you are on a kind of a virtual classroom, sometimes it can be quite hard to maintain attention and exercises don't always go down well. And then. Also, if you're thinking about a digital module, where there's no human interaction, you need to make sure you've got mixed media, like not only your text, but you've got kind of interactive exercises.

You've got pre-work and post-work and different case studies in there. So if. I think giving that variety to really attract all learning types and to really keep the audience engaged, and keep the learning really relevant, is what is what's gonna kind of change things for the better to make learning even more interesting.

Alex: And I'm gonna ask you to do some predictions. Now you've got a lot of, experience in the retail space. How do you think it's gonna change fashion and retail?

Hannah: I think there's a lot of cool stuff that can be done, to reduce returns. So if you can kind of match your makeup online now to see what shade you need to buy or. Kind of see an avatar with your body measurements, and an outfit on that you have seen, to kind of see how it would look on you before you even order it and get it home.

Like that virtual fitting room, vibe as a kind of consumer is really cool. There are already lots of things in place in terms of forecasting based on sales, how. How much and of what size to buy in the future, which leads to less waste, which is really cool. And something that I know is already happening and I know will con like continue to happen and grow, is that rather than kind of doing lots of different samples in lots of different colorways, you'll be able to kind of create those using AI.

So once again, that reduces waste, which is fantastic. Um.

I think kind of the AI models and all the really beautiful things that I see that are kind of like in AI photo shoots, make products so much more compelling to buy. And I just get blown away every time I kind of see the capabilities that are out there. So I guess it will in some ways. Make the online shopping experience more kind of like thrilling and exciting.

As we were talking about that in the classroom or virtual classroom, you know, it used to be that nothing beat going into a store, to buy your clothes because of the store ambience, et cetera. But if you can create that really special space online as well, using AI, that's very personalized. That's definitely where I see it going and my predictions.

Alex: Good, good. Safe predictions are good. I'm gonna move to the quickfire questions now. If you had a magic wand and you could do any job in the world, what would you do?

Hannah: That's so hard to answer. That's so difficult. I mean, it's not very quick fire. Um.

I've always enjoyed creating and I've always enjoyed writing and talking to people. So if it wasn't learning, I guess it would be in kind of like broadcasting or or in some kind of journalism. But yeah, let me think about that a bit more and I'll come back to you.

Alex: Hopefully appearing on here is the

Hannah: I mean, yes, Alex. That wasn't, that wasn't intentional prompts there, but you never know.

Alex: And what's the, what's your, one daily habit that that keeps you effective?

Hannah: I'm a big reader, so I always challenge myself to not doom scroll for 15, 20 minutes and to learn something. Um. And then I also read on the side, like novels or kind of like, factual books. But yeah, I'll read the paper or I will kind of read a paper that I've found that I've saved from somewhere on, something new that's happening or something that's kind of happening in another company.

So, yeah, I'm quite curious and love to learn.

Alex: And what's one thing you do regularly to help protect your mental health?

Hannah: Go for walks. Outside on a day like today where it's nice and sunny, with some tunes on. Can't beat it.

Alex: Good. All right. That's it.

Hannah: Cool. Thank you.

Alex: Thank you very much for joining me today, Hannah.

Hannah: Thank you, Alex.