The Catalyst by Softchoice

Bonus: The Catalyst gets clear on Microsoft Generative AI – Getting ready

• Softchoice

Share episode

'Look before you leap' is often the difference between success and failure when it comes to adopting a new technology.  

When it comes to Microsoft Generative AI, the importance of assessing your organizational readiness before you implement can’t be overstated.

On this episode of this miniseries on The Catalyst podcast, step into the 'Assess' phase of the Copilot Adoption Journey with Matt Vasil, Director of Emerging Technology at Softchoice. We get into the essential preparations for implementing Copilot for Microsoft 365, from measuring the potential licensing impacts to minding security gaps to answering the question, “Are our people truly ready?” 

Matt offers key insights into how to evaluate your environment to make sure your Copilot adoption stays on course and ultimately sticks the landing. 

Featuring: Matt Vasil, Director of Emerging Technology at Softchoice

The Catalyst by Softchoice is the podcast dedicated to exploring the intersection of humans and technology.

Aaron:

Over the last year, we've seen huge anticipation for Microsoft 365 Copilot, and it's finally here. For most organizations, the question isn't whether to adopt, but how and how fast. Early adopters stand to gain the most, but moving ahead without a plan comes at its own risks. You're listening to The Catalyst Gets Clear on Copilot Adoption, a mini series hosted by Softchoice's Braeden Banks. From building a business case to getting security, rollout, and adoption right, we're going to help leaders like you chart your best path forward. It starts right here, right now. On today's episode we ask, how do you assess your Copilot readiness?

Braeden:

The old proverb, measure twice, cut once, applies just as much to charting your path to Copilot adoption as it does to carpentry. Before you put Copilot in their hands, you'll have some questions to answer around the impacts to your licensing and security environment, along with the safety Of course, how you're going to ensure people actually use it. Today, we're talking about how you know you're ready for Copilot with a simple step by step process. I'm talking with Matt Vasil, Director of Emerging Technology at Softchoice, about the importance of making a detailed objective assessment that considers licenses, security, applications, and data before you let your people loose on Copilot. Thanks so much, Matt, for joining us here on the Catalyst podcast, our Copilot Adoption mini series.. So today we're here to talk about specifically the second phase of our, you know, recommended Copilot adoption flight path the assess phase, which is all about getting ready in a technical sense for adopting Copilot. So Matt, I wanted to ask you from your point of view, what are the biggest misconceptions about what it really means to be ready for a Copilot?

Matt:

Yeah, I'd say, you know, the first and foremost one we, we came across was that, you know, licensing readiness equals Copilot readiness. And there's a misconception that just because you may be licensing ready for it, you've got all the underlying licenses that you can flick the switch and start realizing value. I think this is coming from kind of this preconceived idea that most users are just going to get it and be proficient as soon as they get access. Well, it may be true for a small percentage of early adopters and like technically savvy users, but for the large majority, that's just not the case.

Braeden:

Right. And even early on from a perspective of things like security, data, and kind of the governance around all of those things, what's the sense that you have? Like, how bad could it really get if an organization were to say, cut corners or skip ahead on this kind of a readiness assessment? I

Matt:

guess

Braeden:

in

Matt:

terms of how bad it could get, it's probably a range from a best case of a kind of just a wasted investment of time and capital when. You know, folks don't adopt or use or doesn't deploy correctly to probably worst case, which something more significant, like your employees easily accessing and leveraging sensitive information that you previously thought was secure, but copilot makes it much more readily available to folks. So taking that information, creating content from it and potentially disseminating that information.

Braeden:

Right. I mean, even in my own experiences with Copilot, I've had files that I've worked on recommended to me, you know, supporting tasks and questions I've asked for it. And I'm like, I don't know, are those really ready to be shared with other people? So it takes a new perspective of what it is you're creating and sharing. Within that Microsoft like ecosystem and that kind of tees up like what are the areas that might be easy to overlook or assume that it's okay to leave as is. I mean, where are we really seeing organizations like soft choice, you know, risking things by cutting corners?

Matt:

Yeah, I think. You know, there's a phrase that's going around, which is security by obscurity. When you've got a whole bunch of folders and files, you know, hidden in SharePoints, personal users assuming that because they're four layers deep, someone's not going to be able to readily get at it or access it. Just having the, the awareness that Copilot will, will easily surface that information and bring it into documents. I think. One of the biggest things we found in our journey is that we really didn't have the adoption change management skill sets within our I. T. organization to help assist with our rollout. They're so important to a successful rollout, making sure users are proficient and really getting kind of at that business value that you've originally identified. I think probably one of the other areas is that productivity improvements automatically translate into business achievements. So it's easy to go into this thing thinking that, okay, I'm returning 1 to 2 hours a week. And that's a slam dunk business case. But the question that we quickly faced is, how do those extra hours translate into things like business objectives and strategic measurable improvements to help move your business forward?

Braeden:

You know, taking all that in mind, like how in depth can that upfront readiness work actually get? I mean, like you said, there's sometimes this impression that you can just kind of turn it on and it'll start delivering those benefits. But like, realistically, what kind of. Upfront investment and effort is an organization looking at,

Matt:

I think, just generally, the more you invest your time in the planning and the assessment stages, the better off you're going to be to understand and realize the benefits and identifying if you're licensing ready. It's just, like I said, one element of it. Really, this involves. Do you have the data classification policies and permissions in place to protect the sensitive data so that users can access? What they should access versus what they can do. You have their resources in place to help with ongoing technical support, sustain change management adoption efforts. And then I go, you know, always back to the business case. Like, how do you plan to ensure your employees are using it in a way that drove that initial business case in the 1st place? Like, what are the management practices to ensure their effectiveness?

Braeden:

No, that makes a lot of sense. And I mean, you know, do you get the sense that most organizations are mostly ready or are many of them like significantly far behind?

Matt:

What we're seeing is that most aren't ready, meaning that they don't really know what to do. We see a lot of organizations buying 10, 15 licenses and then giving it to folks potentially within the IT group, like my group, to play with. The challenge is that without kind of that preparation, they really don't know how to properly evaluate it. And really to prove out a business case, they don't have awareness of how the tool should be used and what are the outcomes that could be seen, nor do they have awareness of kind of the risk of expanding that beyond 10 to 15 users.

Braeden:

One thing that we recommend as soft choice for customers going through this process is a readiness assessment. Can you talk a little bit about what output our assessment provides and whether those can tell someone about how much work is ahead of them before they're ready for copilot? I

Matt:

mean, assessment or otherwise, it's always starting with the why, right? So what are the business problems the customer is looking to solve? And then how do you turn those into use cases and then assign those use cases to actual pieces of functionality or personas that can apply to their business users? Once you've got that, then I think you can get into the meat of the assessment, which is identifying, really, are you licensing ready? So does the customer meet? Minimum requirements as far as licensing, versioning, and the assessment evaluates readiness against kind of a red, yellow, green stoplight scoring mechanism. We then layer on security reviews to help customer understand if there are compliance, data protection, or data retention concerns they should address. And really, finally, the focus on adoption planning. How can we ensure that all that prep work that went into licensing readiness, security readiness, and that went into developing that business case is actually being realized by ensuring our users are leveraging CoPilot effectively?

Braeden:

No, that's amazing. So Matt, just to take it a little bit of a different direction, in our own CoPilot adoption process in which you were involved as part of the early adopters program with Microsoft, could you tell me one or maybe two things we wish we had known or perhaps done differently from a technical standpoint before we embarked on that process? Or did we kind of nail it? From the get go.

Matt:

I would say we were one of our own first customers to go through the readiness assessment. So, from a technical standpoint, we were aware of all where we stood from a licensing and a data protection perspective. I think if you were to re ask that a bit more generally, I'd probably double down on the effort we put into tying Copilot use cases to clear and measurable business objectives. So now that you've deployed it effectively, you've protected your data and you've got your users using it, what is the value that it's returning back to your organization? That's something clean and easy to measure. You can communicate back to your stakeholders.

Braeden:

Right. That makes a lot of sense. You know, how much does your in your experience that old fashioned resistance to change play a factor in getting ready for copilot? And this can be from a technical team standpoint or, you know, lack of interest in adopting something new or or even just on the people end on the adoption phase. Does that weigh in a lot in your experience?

Matt:

Yeah, absolutely. I mean, with any significant change, change resistance plays a factor, and I'd say kind of any move to kind of AI technologies right now that's almost compounded, which is kind of why when we were getting ready to move and deploy kind of this user study and our rollout, we picked this cross section of early adopters, potential change resistors really have kind of wide perspective so that we could tailor adoption plans to meet both those groups kind of where they were at. One way we looked at it was our general Copilot for web adoption, like how often were our employees accessing these chat GPTS type tools? Because it's a good indicator of, you know, if they're ready for a change or who could potentially be champions in your. Copilot for 365 rollout and just a good indicator of, you know, an AI appetite and acumen across the organization,

Braeden:

right? Because I suppose it does raise the significant risk of that sort of AI powered shadow it adoption. I imagine there's a lot of risk presented by employees or internal users going out and using, you know, the public LLMs or other similar tools. That maybe aren't sanctioned and putting sensitive data out there to get, you know, whatever the return is on it. Like you know, create this report or help me write my executive summary or something like that. Right. Is that part of the process of readiness?

Matt:

Yeah, absolutely. One of the most important kind of milestones we, we stood up was. Producing a responsible AI use policy. Previously, we had some guidance, but not a formal policy. And really, the policy kind of talked exactly what you just mentioned. Like, what type of data? What type of tools should you be using? When it really put a lot of accountability still on. The end user to own the content that's being produced and to use their own judgment to make sure if it's accuracy and correctness before sending that out. But I'd say that's a pivotal piece of kind of any company rollout is make sure they've got a, a use policy around it that supports the behavior that they're trying to encourage.

Braeden:

And in kind of preparing for the conversation to, you know, I kept coming across the phrase of garbage in garbage out. And that kind of made me think of the specific. Context of data readiness and I kind of wanted to ask you a little bit about how soft voice approached making sure our data to be connected with copilot within Microsoft 365. How did we approach that? And how do you know when it is clean enough? Is there a way to know?

Matt:

I don't know if. You can never fully know based off of the, just the general limits or lack of limits in the copilot ecosystem. Like everything in your tenant is up for grabs. I think where you need to put focus is really understanding what's sensitive. What is off limits for folks? Where. Should they play and where shouldn't they play and having kind of those retention and data labels around it to ensure that can't get out. But it does come with a good amount of business process change as well. It's not something that you can just stand up a data label called sensitive and then expect folks to use it. Correctly. So coming with that kind of guidance and saying, Hey, we've stood up this new data protection policy. Here is the process to classify a piece of data correctly. And here's what's going to happen if you do. Or more importantly, what's what's not going to happen. So I'd say that's kind of where the journey we went through is taking a look at our data and focusing on the personal information, highly sensitive information first as a starting point.

Braeden:

Right. And it kind of brings back that that age old chestnut in the I. T. world of like your biggest threat is, of course, the human breathing 12 inches from the screen, right? Like, no matter how technically ready you may feel you are without that human guidance and business process change. I mean, you're still, I think, considerable risk of things going wrong. So that's got to be a two pronged approach. Agreed. I'd say the same thing. So before we wrap the conversation, having gone through the experience of rolling out Copilot at SoftChoice and, you know, we're not to toot our own horns, you know, we're probably immersed in this sort of generative AI space in the enterprise tech world here at SoftChoice for those out there who maybe have a workforce that's not necessarily as advanced in this kind of area, knowledge wise, is there anything you'd want them to think about as they kind of approach this could be daunting project?

Matt:

Yeah, I think it's it's really understanding. What you're trying to get out of it. First, understand the use cases, potential benefits. I think the more you can get your employee base immersed in, you know, just general prompt understanding, the better off you'll be. And, you know, that could be through, you know, copilot for web activity, something that most have available to them. Standing up adoption or training or drop in sessions on how to use it, get folks really familiar with that prompt engineering perspective, which translates so well into copilot for 3 65. And if you've got a user base that's familiar with that. The adoption activities will

Braeden:

will be much less. Any last thoughts for our listeners before they move on to episode three?

Matt:

Yeah, I think, you know, once you've gotten through kind of the plan stage and you're into assess, it really is around, you know, what does readiness look like for your specific organization? Let us help kind of identify potential risks and challenges. both technically and organizationally, and make sure you're really aware of kind of the financial and human resources required. That's how I would kind of encompass the SS stage. Thank you very much for the time.

Braeden:

And thank you so much for lending us your time and joining us here on the Catalyst podcast. Thanks so much, Matt. Thank you. While the pressure from all sides to get going with Copilot is palpable, you cannot overstate the importance of being sure your data is clean, your licenses are in order, and your people are ready to unleash their potential with this technology. Thank you. By properly assessing your readiness, you can prepare and protect your data and set your organization up for success. That's all for now, but the journey doesn't end here. Join us in the next episode, where we'll explore another golden rule of copilot adoption, start small. Until next time, I'm your host, Braden Banks. Thank you for listening.

Aaron:

Imagine a world where your organization doesn't just follow AI trends, it leads the way. Softchoice is ready to help you build that world with Copilot for Microsoft 365. Why Softchoice? We help our customers navigate the whole copilot journey with a proven consulting framework, secure implementation experience, and deep certification in Microsoft technology. We know Microsoft better than anyone. Every copilot needs a navigator. Softchoice is yours. Visit softchoice. com slash Microsoft dash copilot to learn how we can help you unleash the potential in your people and technology with copilot.