The BCN Podcast

How to get the most out of Microsoft Copilot: Adoption tips for businesses

• BCN

Share episode

In this episode of the BCN Podcast, we explore the approaches businesses should take when implementing AI tools, focusing on education, security, and effective adoption strategies. Experts Fraser Dear and Johan Venables share insights on how to identify business cases, address data security concerns, and ensure successful AI integration.

• Starting with education to help staff understand AI capabilities and address misconceptions
• Identifying the right business use cases for AI implementation, including individual productivity tools versus process-focused solutions
• Conducting thorough data discovery and security audits before implementing AI tools
• Building appropriate governance frameworks and AI councils to monitor and control AI use
• Creating supportive communities and identifying champions to drive successful adoption
• Using Microsoft Viva Insights to measure productivity gains and demonstrate ROI
• Recognising that AI adoption is a continuous journey requiring ongoing support and education

Visit our website at bcn.co.uk to learn more about how we can help your business begin or improve your AI adoption journey.

Peter Filitz:

Hello and welcome to the BCN podcast. Today we'll be talking all things AI, adoption and the approach we're seeing businesses take, what works, what doesn't work, what they need to look at and how they can get started. Once again, I have the privilege of Fraser Deer and Johan Venables to join me, peter Phillips, your host, where we will be exploring this in more detail. Gentlemen, why don't you give our audience a quick introduction into who you are, in case they've missed the previous episodes? Fraser, you first.

Fraser Dear:

Hi everyone, my name is Fraser Dear. I look after all things AI and data innovation here at BCN, so I have to say at the moment, it's never a dull day.

Peter Filitz:

Excellent, yeah, fair to say, so much change, right, fraser Keeps us interested, that's for sure. Johan, over to you, hello there.

Johan Venables:

I'm a cloud consultant. I help and support our clients through their digital transformation journey, focused heavily on the Microsoft 365 stack, especially productivity tools and security.

Peter Filitz:

Thanks Jan, Thanks Fraser. As always, a pleasure having you. So lots of talk around AI. I know every other podcast we're doing is around AI. It's topical. Businesses are also interested in understanding how this emerging technology can benefit them, their staff and the way in which they run their businesses. People want to do more with less, be more efficient, save time. There are still some businesses that haven't started this journey, or businesses that are really keen to get going, but not quite sure where to start. So I guess it's about laying the groundwork, and what I'm quite keen to understand from yourselves, guys, is what are the first steps businesses should take before implementing AI tools like Microsoft Copilot, for example. Johan, I know you've been heavily involved with the whole Copilot launch and the adoption approach here at BCN. Do you want to take that one first?

Johan Venables:

Yeah, absolutely so. When I launched Copilot, there was a lot of fear and pushback on getting people to test it, because Microsoft said go and nominate a few key staff members, especially your productivity workers, who generate a lot of content, read a lot of emails, et cetera. Get them to have a play with co-pilot and document some use cases and review it and then see how everyone else can use it. The biggest fear people had was data exposure because you know, with SharePoint it's quite a common theme where people overshare information, so that slowed down the whole adoption of Copilot. So what we've done is we've developed a webinar where we can basically inspire staff to use Copilot Chat, which was the free version of Copilot, so that they can get their journey started, where you could upload a document, summarize a document, you can help it rewrite content, but you had to obviously copy and paste the content between the two apps. But that was a good starting point. So we'll do a webinar to inspire them, show them some use cases, do a few live demos and also we would run through with them.

Johan Venables:

On safe use of AI be careful of third-party AI tools. We explain the risks. We talk about AI being biased. You need to make sure you double-check all the information it provides and we remind people it's a tool there to assist you not to do everything, so you need to obviously spot-check all of that. And then, as part of the webinar, we also show them the full capabilities of Copilot 365 so that we can get some ideas flowing, and we would challenge everyone to start using Copilot chat to get familiar with it, but at the same time, ask them to write some good use cases for it so that when the organization has sorted out their data, the security around their data they are good. The security around their data, they are good to go and start assigning licenses to these sort of early adopters.

Peter Filitz:

So I think it's fair to say, from what you've just explained there, it starts with the educational piece. Right, absolutely Educate staff, business owners, stakeholders around what we mean by AI in the workplace today, because I think the term AI is thrown around a lot. There's a lot of misconception around what we're talking about in the use cases in today's business space. So it starts with education and then with that you can then inspire staff, key decision makers, stakeholders, to then explore business use cases. And these business use cases I mean, what would you say Fraser is a sensible approach, because businesses often say, well, I've got so many, how do I choose? What is the approach you've seen when it comes to that that works?

Fraser Dear:

Yeah, it's a really good kind of place to start, because for anybody who's heard me talk about AI over the last two years or three years, the word AI is somewhat meaningless because there's so many parts to it. You know, there's so much of it that AI is no longer enough. It's well, what part of AI are we talking about? And I think, for businesses kind of starting now, there's kind of in my head at least two modes. There's buy a license for enablement, and this is when we're talking about Microsoft's M365 co-pilot. That's what I'm talking about. You don't need to do a massive digital transformation journey. You buy a license for maybe a pilot group or a small team that you're going to invest in and adopt in and so they can start to elicit those user stories for your business and demonstrate return. You know, actually, if I did this before manually, it would take this amount of time, whereas if I did it with co-pilot, it takes this amount of time, so you can actually demonstrate there is a benefit from that license approach.

Fraser Dear:

Copilot license is great for self-productivity.

Fraser Dear:

I want to improve my activities in my daily life and perhaps those around me, those consuming the outputs of my work, but actually usually businesses have got a business system or a tool that they're using, or a process, and keying those processes today into Copilot isn't quite so straightforward for the general business user, and that's where we would start to think okay, right, if you want to adopt AI tooling into your business process, let's be really clear about what it is you want, what the outcome you're looking for is and how that AI is going to be applied to that process or service.

Fraser Dear:

So it's identifying that low-hanging fruit, that really big, mundane, repetitive task, bottleneck in your business, and applying AI processes or tooling around that process to enhance and make that return, make that demonstrable activity, Because once you've got one that works, there's going to be a whole backlog of oh well, what about that one, and what about that one and what about that one? And so then you can start in a controlled and stable way, start to roll out and build your AI whether it be ecosystem platform, whatever it may be, whatever those tools are within your organization or those bottlenecks to then start slowly but in a controlled and returned way. That's what I would suggest.

Peter Filitz:

Great yeah, and that's saying so true as well, not all AI is equal. So it's making sure that you identify the business case and then work to identify the right AI service and solution for that particular business use case. The way I like to sort of explain it to people is that you've got AI personal assistants, which are obviously aimed at driving individual user productivity and efficiency, and then you've got the departmental-wide, organizational-wide AI assistants, which are aimed at changing and improving business processes that touch multiple people or multiple departments within the business. But I think it's fair to say, with all AI opportunities and journeys, it does all start with the data right and making sure that what we overlay the AI on is a single source of truth, it is of a good quality and, more importantly, it is appropriately secured. Johan, do you want to talk us through the approach that we take when it comes to looking at data security and governance for businesses?

Johan Venables:

Yeah, so typically the first thing we'll look at and see if you can get funding support from Microsoft around data security. So if you qualify, that's great, microsoft will fund the whole project. If not, we can have conversations around there to see how we can help. But our typical approach is really to do the first place we start at is to do an AI discovery. So we first want to understand what AI tools people are using outside of the Microsoft ecosystem and we want to build a list of AI tools people are using and we also want to see what data they are using with these tools. And then from that we'll work with our customers to build a policy to either control what data they're allowed to use and not allowed to use with third-party AI tools, or it could be a case where they go no, we want to immediately block it.

Johan Venables:

It just depends on each and every business. Everyone's got different requirements. We've got a lot of companies that's a huge fan of ChatGVT and they still want to continue to use it alongside with Microsoft Co-Pilot, and then we'll just basically work with the customers to put conditions in place on what it is they can do with those tools. At the same time, we'll do a data discovery because we need to identify what sort of content they have, whether it's in SharePoint, business applications, on a file server, sql. We will pretty much do an audit and scan for sensitive content and then map out where those sensitive content lives around the data sources so that we can then from that, build a plan on what are we going to do with that data.

Johan Venables:

What do you classify as very sensitive, too medium, too low? And then we'll build conditional access policies based on the severity of the data and control it in a way where people can't do things like share it externally or download it or use it with third-party AI tools or whatever it may be. We'll put strict conditional access policies and access control around that based on the information we found in the audit. Because, look, we can go on and deploy all the standard policies and create a baseline for you, but the best way to approach data security and understanding what people are doing with AI is to do that discovery piece first, because once you understand where they are at from a data security perspective, that's when you can start making recommendations excellent.

Peter Filitz:

Yeah, am I right in saying yeah? And it's purview that we use to help businesses undertake that security audit, and we did a podcast series on purview and what it brings. So if you missed that, be sure to check that out. And then, in terms of governance, ethics and risk management, I know we're also big advocates of setting up AI councils within the business right, because it's important that any use of AI within organizations is closely monitored, managed and controlled. Do you want to talk us through what we're suggesting businesses look at setting up on that front?

Fraser Dear:

Okay, so you know, let's go back to what I said earlier on We've got two sides. So, if we're thinking about Copilot, microsoft have provided quite a lot of information available in the Microsoft portals and the administrative portals to look at kind of user engagement, which users are active and which tools that they are active within within the bounds of the M365 Copilot license. Now, in order to kind of dig a little bit deeper, that's when you need to start to look at your audit logs and start to understand kind of what people are prompting against. But actually there's quite a large amount of governance and control around M365 Copilot. If you're thinking about going down the other route, which is about business process enhancement using AI, using AI tools, using advanced AI, using, you know, pro code routes, if you like, creating something bespoke for your business, that's where you actually build that control into the process that you're actually using.

Fraser Dear:

Again, I mentioned kind of red teaming on a previous webinar. At the end of the day, your business users are the ones that will be able to provide that dynamic governance. That is this doing what we want it to do. And what I would say on that side is, of course, we can build physical controls around it and we can build security controls around it, but the reality is you wouldn't employ someone and just say get on with it and never check up on them. This is where you need to apply the same rules as you would to any other business process here within an AI environment Audit it, check it. Is it still giving you the results you're expecting to get? Those controls are actually just as important as ensuring that it's got access to the right data, ensuring that it's not been overshared, ensuring that it's not got access to data that it shouldn't have.

Peter Filitz:

Excellent. Thanks, fraser, for that. I think that brings us close to the end of today's conversation around the adoption. I just wanted to get any words of wisdom, so to speak, from our two experts today around what businesses should consider if they are about to embark on this journey or if they've started that already, fraser.

Fraser Dear:

Look, ai is an emotive subject for a lot of people within businesses. It's a change, and it's not just a change in the way your business works. It's a fundamental shift in the way that businesses are working. Think back to when we didn't have tech and the fax machine came along and people were like, oh my God, this is incredible. And some people were like, oh my goodness, I'm not going to have a job anymore because I'm not going to be writing things down no-transcript.

Peter Filitz:

And that's why it's so important that education and awareness is built and really forms the foundation of any AI adoption approach, Because for any project or change to succeed, you need the buy-in from the people who are going to be using it and therefore it all comes down to the way in which it's positioned to them and adopted by them ultimately, I mean don't get me wrong, this is a financial transaction, this is an investment, so there's a whole business aspect of it.

Fraser Dear:

An investment, so there's a whole business aspect of it. But just don't forget your end user in all of this, because they're the ones that will be anxious about the adoption, about the forthcoming new MP65 copilot licenses that are going to land on your desk next week, because of course, they don't know what it means. So make sure they're involved in the conversation, absolutely.

Peter Filitz:

Johan, parting words from you.

Johan Venables:

Yeah, and I can echo on that, because the thing is, it's important you know to be honest with your staff that you are embarking on a journey to adopt AI with the organization. But be clear what it is, your plans are, what is the outcome you're looking for? And you know, communicate to them and say to them this is going to make your life easier so that you can do more. Get them excited about it, because, yes, a lot of people are concerned about AI and every conversation I have. They tend to joke a little bit in the conversations by saying, oh, is it going to take my job? And I immediately go don't worry about it. That's not the case. Right, it's there to empower you to do better things. But the most important thing I want to share here is that it's a never ending journey, so you're going to have to monitor their activities. You're going to have to make sure they get the support, the training that they need for a successful adoption.

Johan Venables:

Microsoft is launching Viva Insights, a new sort of experience where you, as an admin, can see and understand the challenges people have with adoption. It will give you some recommendations, some tips and tricks on what it is you can do to support them with that. But you can then measure how much time they have saved working in Word, on emails, in PowerPoint, in meetings et cetera, so that you can see that return on investment. And there's always new features coming out with Copilot. I mean when I catch up with clients, which is on a monthly basis. I've got a new feature release to talk to them about every time when I catch up with them. So it's making staff aware. It's building a Viva community, you know, a community where everyone can share their success stories, their problems, their challenges and finding your sort of champions internally to help and drive and support that. Adoption is absolutely crucial for people to get the most out of CoPilot.

Peter Filitz:

Excellent. Thanks, gentlemen, for those parting words. So if you're a business owner looking to embark on that journey, or you've started it and it's not quite going the way you had thought or hoped, you know BCN are here to help you. We've helped hundreds of businesses start this journey in various shapes and forms, and we'd be more than happy to help you. Feel free to visit our website, bcncouk. There you'll find a wealth of information around the products and services that we provide. It's fair to say, AI is here. It's yet to stay. It's very much going to be an integral part of our future and therefore, if you've not started your journey, it is something you really really need to look at getting going with, because if you're not, you can be sure your competitors are Well, we hope you guys have enjoyed today's conversation. Look forward to having you on the next one. Thanks, gentlemen, for your time. Great to see you again. Thanks so much everyone.

Johan Venables:

Take care, bye-bye.