The Copilot Briefing
The Copilot Briefing is the podcast for business leaders navigating Microsoft Copilot, AI agents, and the future of work. Each episode cuts through the hype to focus on what actually matters: strategy, governance, adoption, risk, ROI, and business outcomes. Hosted by Microsoft MVP and AI consultant Steve Corey, The Copilot Briefing helps executives and decision-makers make smarter AI investments, avoid costly mistakes, and turn Microsoft's AI ecosystem into measurable business value. Smarter AI decisions. Better business outcomes.
The Copilot Briefing
Measuring the Business Value of Copilot
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Microsoft 365 Copilot may be easy to demonstrate, but determining whether it is a worthwhile business investment requires more than tracking licenses and usage. Steve Corey explains how to evaluate Copilot through role-specific use cases, before-and-after process measurements, data readiness, security controls, ongoing support, and documented business outcomes.
Topics Covered
• Why polished AI demonstrations do not prove value
• The difference between access, adoption, activity, and outcomes
• Measuring processes before and after introducing AI
• Converting saved time into additional business capacity
• Overshared content and broad Microsoft 365 permissions
• Assigning licenses according to job roles and tasks
• The importance of task frequency in ROI calculations
• Ongoing user support and champions programs
• Documenting evidence before expanding a rollout
• Search, meeting follow-ups, document drafting, and agents as use cases
Key Takeaways
• A Copilot rollout should begin with business processes and measurable outcomes.
• Usage percentages should not be treated as proof that work has improved.
• Saved time creates value only when the reclaimed capacity is put to productive use.
• Permissions and sensitive-content controls should be addressed as part of readiness.
• License allocation should reflect a role’s genuine opportunities to benefit.
• Before-and-after evidence supports future licensing and expansion decisions.
Chapters
00:00 A Business-Focused Copilot Briefing
00:49 The Investment Question
02:02 What Determines Copilot Value
02:54 Meet Steve Corey
03:53 More Than a Software Purchase
05:16 Transform and Measure Processes
06:39 What Happens to Saved Time?
08:20 Access, Adoption, and Outcomes
10:27 Returning Time to the Business
11:39 The Oversharing Risk
14:27 Matching Licenses to Job Roles
15:53 Cost and Task Frequency
17:02 Ongoing Support and New Features
19:09 A Fiscally Responsible Approach
20:26 Documenting Measurable Outcomes
22:09 Practical Copilot Use Cases
23:32 Proving the Before and After
24:54 Closing and Future Episodes
⌚ Book your free on 1 on 1 call with me to discuss your business challenges and how AI can help 👉 https://go.stevecorey.com/copilotbriefing
🔴 Need help with Microsoft 365 Copilot adoption, governance, and employee training? Let's build your strategy ➡️ https://go.stevecorey.com/copilotbriefing
A Business-Focused Copilot Briefing
SPEAKER_00Welcome to the first real unique episode of the Copilot Briefing. In this podcast, I'm going to be really aiming everything towards you, a business leader. I'm not going to be using all the technical terms I might be using over on the YouTube channel that would be targeting more of the implementers. This is specifically business focused, and I'm focusing on things like the return on investment, the strategy behind a lot of the reasoning I use as a consultant when I'm working with my clients. And the first topic is probably one of the most important, and also a topic I'm going to be touching a lot on at Microsoft Ignite during my lightning talk. So if you are going to be there, definitely check that out and look for me on the floor. I'll be definitely looking forward to talking with you. Microsoft
The Investment Question
SPEAKER_00365 Co-Pilot is it's easy to demonstrate. Of course, all of the AI products have flashy demos. So it's easy to see how good this thing is, but it's a lot harder to evaluate as a business investment. A polished demo just shows you what the technology does, but it can't really tell you whether your organization's actually ready, which employees should receive that AI tool, or whether the value you're going to be getting will actually justify the cost. In the case of M365 Copilot, this could easily cost you hundreds of thousands of dollars a year. And I see a lot of customers making the same mistakes over and over again. And I'm going to be touching on a lot of those in this episode. So in this first episode of the Copilot briefing, we're going to answer the question that matters to executives. Is M365 Copilot worth the investment for your organization? And unfortunately, the answer is not necessarily a universal yes
What Determines Copilot Value
SPEAKER_00or no. It really depends on where the work is, whether your Microsoft 365 environment can support it, whether your data is ready for it, whether anyone's actually accountable for turning usage into value. That's the big difference I'm seeing between successful and unsuccessful implementations. I see customers trying to go their own route with Copilot, and they see some metrics that, but and the numbers look good, but they can't quite justify or prove whether there's actually any value being generated by that large expense. And so they start to turn to, okay, let's let's go with Claude. Let's you do Claude Enterprise. They're going to run into the exact same situation in the same problems because they're not trying to fix the root cause that we're touching on in this episode.
Meet Steve Corey
SPEAKER_00If we're just now meeting, my name is Steve Corey, and I'm a Microsoft MVP and AI consultant helping organizations turn Microsoft's AI ecosystem into measurable business value. My specialty is Microsoft 365 Copilot, but also agents. I do a lot of developing of agents and building solutions using various Microsoft AI tools, not just copilot, because I think that the right tool always needs to be picked for the right job. You start with the problem, though. And so sometimes co-pilot is the right tool to fix a particular problem. Sometimes it's not. And I have no issues telling a customer that.
Transform and Measure Processes
SPEAKER_00So while someone might be experimenting and they're they're having co-pilot an outlook, summarize an email, they're they're reading an email faster. Okay, that's great. They're saving some time. But what really should be happening is someone should be looking at their processes, their day-to-day processes, what things take up a lot of time, or what things are highly repetitive, what things could be improved with some sort of an AI transformation. And then intentionally measuring the time it takes them today without AI to do that process, especially at scale, and then introducing AI, changing that process to leverage AI, whether that's copilot, whether that's an agent, something along those lines. You're leveraging AI, and once you've got that process correctly transformed, remeasure that process. How much time does it now take? Sometimes it actually might take more. That would indicate you probably don't have the correct transformation. You're not quite leveraging the AI in the right way. But in a lot of times, you do see a lot of time savings there.
What Happens to Saved Time?
SPEAKER_00But that introduces a different problem. What are they going to be doing with that time now? If you free up two hours a day of someone's of someone's day, of the work they're doing, but then that two hours of savings, they're just spending it at the coffee station getting some more coffee or chatting it up with co-workers. You didn't actually get anything of value from that person's AI, their their now their new AI process. Because yeah, they could get their work done faster, but the point of introducing co-pilot and AI to their work was so that they can get more work done. Now, if they get 10 hours of work done in an eight-hour day, now that is the measure of success. And if they are even more so, if they've learned how to look at their processes and the processes around them, and identified more and more use cases where copilot can speed things up, and then eventually where agents can now speed things up or automate processes, reduce manual work. That's when you really start to scale that value through the roof. That's how people are winning today with Copilot. It's not just assigning a license and issuing training, and that's it. It does need to be a bigger hands-on, in-depth look. So that's one of the first points I wanted to make is just kind of level setting with what people are doing right and doing wrong out in the world that
Access, Adoption, and Outcomes
SPEAKER_00I'm seeing. I work with clients every single day. And I was talking to one just the other day, and they're evaluating Claude Enterprise alongside Copilot. And this particular client, they kind of like to do their own thing. They've got some in-house experts. And so I advise as much as I possibly can, but in the end, they're going to do things their way. And I'm just going to be there when they can't seem to get any sort of an AI tool there. And I'll be like, okay, are you ready to have a discussion about what you should be doing, why everything has been failing, and what you should do instead. Because buying a license, whether that's Microsoft 365 Copilot, whether that's Bot Enterprise or Chat GPT, whatever their enterprise one is, buying that license is just giving someone access to a tool. It's not creating adoption. Adoption creates activity, but even then, not necessarily outcomes. Just because someone is using Copilot after they've gotten their license, which is pretty common. And it's it's fairly easy to start to get that adoption percentage number you see in the dashboard. It's easy to get that to 85 or more. A lot of people, when they first get AI, they're excited. They're using it for a lot of stuff. A lot of times that just ends up being a search engine. They're talking to Copilot Chat the way they would Chat GPT, uh, some of their other personal consumer type AI products, like the Claude personal one, and they're just using it to ask questions, maybe to find information and SharePoint. And it'll work. It's still not the best tool. They may not be getting a lot of value out of that, but it's going to show up in that percentage. It's going to show up as, oh, look, we've got 90% adoption. And customers will think that that's a success. They think that great, our rollout is successful. And the reality is that can't be further from the truth. Because you have to look at how the work has changed. That's where you see value. And when you see time savings, when you see processes adjusted
Returning Time to the Business
SPEAKER_00to incorporate AI, again, it's what are they doing with the extra time? That has to go back into the business. That needs to go into more strategic work with sales teams. Maybe they've uh really tightened up their process of updating CRM, updating sales opportunities, that sort of thing. And so it gives them a lot of time to really, really work on those emails to your leads to craft that thing perfectly, something they may not have been able to do at all before because they just had too much work they had to get done. So this savings can really put them in a position to be much more successful because they can focus on those things that matter the most. Not updating CRM, but crafting an email that's going to get someone to say yes to an offer. So just remember that time saved is not automatically money saved. Another big issue that is ongoing. Um, most of the time, customers don't ever
The Oversharing Risk
SPEAKER_00think of this one thing when it comes to being ready for copilot. They'll roll it out, they'll still see a problem, but they still in a lot of cases, they still won't actually intentionally stop what they're doing, fix the problem, and then continue on. And that is broad permissions inside Microsoft 365 Copilot. Whether that's in OneDrive, in Teams, in SharePoint, things have been overly shared. It was far too easy for someone to share a uh maybe a sensitive HR document. And when they create the sharing link, they just said, okay, everyone in the organization, because I'm only going to send this link to my own team. So everything's all good. Yeah, until you introduce Copilot, and Copilot is incredibly good at finding information. If you have permission to a document, Copilot can find it really, really easily and quickly. So all of a sudden, employees are finding spreadsheets with everyone's salaries. They're finding job offer letters, they're finding a lot of information they really should not find. And I know you've heard of these horror stories. I hear them on a regular basis. I've heard them for years. And it's because someone was so excited about adding AI into an organization and how easy it is. You buy the licenses, you assign them, there's training you can get from Microsoft, just have everyone start going through the training, and you call it a win. I mean, we already know that's not a win condition right there. There's there is a lot more. But if you've got overly shared permissions or if you have overshared content, then you're going to introduce a lot of risk. The risk actually goes up incredibly high until you're able to fix that. The risk will stay high until you find all of those uh overshared documents and address them at the file level, the library level, wherever that file is, you your SharePoint admins, your compliance people have to work together to make sure your documents, if it's confidential, there has to be a label indicating confidential. All your sensitive content has to be labeled, permissions have to get taken care of. Otherwise, you could end up in the news. And that's not a good thing to do. Um, not in this sense, about this particular situation. Another thing that is good to call out as well, if you're in the mode where you're going to start assigning
Matching Licenses to Job Roles
SPEAKER_00licenses out, understanding that different job roles are going to have different levels of opportunity with co-pilots. Someone who's working with information like documents or spreadsheets, that's all they do day in and day out. Massive opportunities to use AI to improve that process. Compare that to someone who's maybe working in SAP or working with uh uh in-house software where it's a lot more specialized software work. Uh, maybe it's data entry uh into a particular uh custom system, Oracle, perhaps, something like that. That's not gonna have quite the same type of benefit. So different roles may not have much of an advantage at all to using copilot. Frontline workers are another example where it might be difficult to find what process could copilot fit in. It's why I always start with uh the job role, the persona, breaking up all of your employees into personas and roles, understanding what job tasks they perform, and then evaluating that to see is there enough opportunity there to justify that license. $30 per user per month doesn't seem like a huge thing until you really think about the fact that that's three was it,
Cost and Task Frequency
SPEAKER_00$360 a year. I'm not uh uh good at doing math on the spot or anything, but yeah, like it's a decent chunk. It doesn't take a lot of opportunities like what we've been talking about to make up for that license cost. And it doesn't make take much to all of a sudden now you're creating value, you're getting your money back and then some in terms of value with Copilot, but you do have to understand that not everyone will have the same types of cost savings. Another thing is when you're looking at those processes that each different role has, how frequently do these employees perform that task? Because whatever your cost savings ends up being, well, that's multiplied by the amount of times they do that. If they do it one time a month, yeah, it's not a big value there. If they're doing it 10 times a day, that's a very different story. So as you're looking at the team that is managing your co-pilot rollout, your champions program,
Ongoing Support and New Features
SPEAKER_00and so on, you should be making sure that they have those things in mind, that this is not just assigning a license and calling it a day. Uh, users will need ongoing support. They will need a lot of support to help them look at their processes, to answer questions, to constantly show them some of these new tools that continue to come out, because Copilot has got a lot of different pieces now. There's co-work that will probably have an entire new episode on how co-work can change some of these processes because it has its own very, very unique cases with repetitive work. But there's a lot of things, and I see a lot of companies that will just turn off everything new and they stick to just one or two different pieces of Microsoft 365 Copilot. And what they're really doing is they're saying, Hey, I don't care about any potential extra savings. I just want to turn this thing off and not have to worry about it. Sure. I guess that is a fair argument if you don't have the time to investigate what this feature is, what value it brings, what opportunities are now on the table for getting more value. But you've got to remember that there's those people that may not have been a good fit for copilot before, but all of a sudden, oh, this new tool, okay, this actually could help them out a lot. So it is worth taking some time, making sure your co-pilot leadership is looking at some of these opportunities, reevaluating all of the information they've gathered on the roles throughout the organization, what tasks are performed, and hunting for that extra bit of value. I've always been, uh I think the word would be fiscally responsible.
A Fiscally Responsible Approach
SPEAKER_00Um, I've I I'm I'm I'm usually pretty tight with my money. And I want to make sure that if I ever buy something, I I know I can get my money's worth, and then some. I don't want to buy something. If there's the risk, I may not use it. And I've certainly bought some gym equipment um over the past couple of decades, and I think you could agree. Yes, some of that stuff ends up collecting dust, and that just eats me up. Uh I I just don't I I yeah, uh I've learned to temper my expectations uh with the excitement I have for wiping that card and buying that new thing to make sure I'm actually able to get that value. Because once the money's spent, it's spent. Once you uh give Microsoft that large chunk of money for those licenses, it's theirs. And it's on you and your team to make sure that what they're rolling out is going to impact the business in a positive way and not just another cost that shows up on your PL statement. And that brings me back to a point that I've made before is making sure that when you're implementing
Documenting Measurable Outcomes
SPEAKER_00these tools, you have a way to measure the outcome. If you don't start by measuring the length that a process takes, the cost of that process, figure it out in dollar amounts. What does it cost the business for this person to spend 60 minutes to do this one task? Then you introduce AI, remeasure it. Once you've gotten that process transformed, once AI is now a part, the process has changed, remeasure that. How much time is that now? And make sure that's documented. I think that has to be documented, right? You need to be able to show proof that this big copilot investment is doing something positive for the business. Otherwise, you could have a very hard argument with when accounting comes or auditors come. People are asking questions about whether this was good, whether, like, should we, and more importantly, should we buy more licensing? Or should we hold off on any? Should we buy more? What is going on with that co pilot program? And are we ready to grow it? Because once you get up to half a million dollars a year in licensing or more, it it it becomes a bigger risk, a bigger opportunity. And it all depends on how your copilot team is managing the changes that copilot should be introducing. Now, some of the things that copilot really, really is good doing that are easy use cases
Practical Copilot Use Cases
SPEAKER_00that you could fit into different different roles. Reducing the time, finding information. That's an easy one. In fact, there's copilot search. It's just a search engine, but it's all using AI. There's uh meeting follow-ups that copilot can help some, you know, quickly generate those follow-ups. Uh shortening uh drafting a document, the amount of time that takes, revising that, iterating over a document. GoPilot's it's freaking incredible at that stuff. Allowing teams to handle a lot more work without having head count. That's where you see some of these specialized agents, like the facilitator agent, other things that are just part of that license you're buying. And hopefully that is part of the adoption work, part of the training, part of that work you're doing with teams to help them understand what these different pieces of the puzzle can do for them. And little by little, they're adding more and more and more AI things. They're getting a lot more value from that $30 license. Those are some of the really good use cases that will generally work across the board, with probably with the exception of a frontline worker. So when you're looking at some of those outcomes of the ways that copilot could really help, you should be looking to
Proving the Before and After
SPEAKER_00and asking yourself, which ones would be how be valuable enough to justify that investment? And how would you prove that it happened? Always be thinking about that, proving the before and the after of what copilot did for a particular process to help an outcome. And sometimes it can be very difficult to look at what processes you have across all of your roles and developing those use cases, working all these workshops and coming up with what use cases would be appropriate or not appropriate for co-pilots so you don't waste your time. So your team doesn't waste their time. These are some of the many things that I do when I'm working with customers on fixing oversharing or developing custom agents to satisfy particular use case requirements, making sure that people are able to get the most value they can get from that existing license. We're not even necessarily talking custom agents and copilot credits and some of that kind of stuff, but just that existing investment, getting the most value out of that. Definitely reach out, uh send me an email. Uh, see some contact info in the show notes if you want my help
Closing and Future Episodes
SPEAKER_00to get some of this stuff done, to work hand in hand with your champions team, with your copilot leadership on this sort of stuff. Now, in upcoming episodes, we're going to be touching on a lot of other aspects of copilot measuring ROI and adoption and problems that I see. I know we've touched on some problems already in this episode, but touching on a lot of the other problems that I see, uh, some of the war stories that I think will be really entertaining and useful for you so that your organization can avoid some of these mistakes. And boy, let me just tell you right now, there's some big ones. There's some politics, there's there's a lot of uh there's a lot of strong personalities that could come into play for some of these things. So we're gonna be talking through all of that and hope you found a lot of value out of this episode. I hope that this has started to give you some clarity on how to properly evaluate your own co-pilot program. Feel free to reach out if you want to discuss anything. And hopefully, I'll look forward to seeing you at Ignite if you're going to be in San Francisco this year, because I am looking forward to having a lot of really good, valuable conversations. See you on the next one.