The Playbook for Enterprise AI Adoption with Walt Wear

Chris Nichols (00:09.304)
Walt Ware, welcome to the show. Wanted to have you on. I am one of your big fanboys out there. hang on your every word. Walt, you have been instrumental at helping us our organization adapt to copilot and basically humanize AI and make us all more productive every time I listen to one of your shows that we’ll talk about. I pick up a couple extra trip tips.

Every time I go to AI Central, our central repository, I pick up some tips and you’ve you’ve really helped make me more productive. So welcome. Let’s talk about. Let’s get into it.

Walt Wear (00:48.338)
Yeah, sounds great, Chris. Thank you. You’re too kind, first of all. but I’m glad that I’m able to add anything, especially to especially to your group and to our bankers across the organization. So thank you so much for that feedback.

Chris Nichols (01:03.81)
So real quick, just for the listeners, kinda tell us where we are on the journey at South State in terms of ad adoption of Copilot, how you started how the idea came about and what we’ve done so far.

Walt Wear (01:15.934)
Yeah, absolutely. So where we are with our copilot adoption, I would say that, you know, I want to separate Copilot into its two different buckets here. You’ve got the features that come basic with your premium licenses as opposed to some of the other f advanced features like Studio, where you’ve got to pay additional fees for those those features. When we look at just that basic license and what that affords our team.

We are on a really strong path to adoption across the organization. So earlier this year we started with only 300 people who were licensed with our premium license. We are now up to, I believe the latest measurement I saw that was official was 1800. I believe we’re actually past that now by maybe even another thousand.

But off that 1800 number, I can give you a percentage of our our adoption rate, and that was sitting at ninety-five percent. So that is people who are using, actively using Copilot Premium in its capacities. we measure this against Microsoft’s internal target. They want to see their users, their their partners at an 85% or better adoption rate. So 10% above that.

We’re doing really well. We have hundreds of people attending our office hours every week. We’ve got hundreds of visitors to our AI Central site every week. People are very engaged with this. And when we look at our total prompts, I think we’re somewhere near 200,000 a month at this point. So you’re talking dozens of prompts per person on average, which is really what we want to see. We want to see people reaching for this multiple days, multiple times throughout their day.

and we’re getting there. We are really starting to achieve what we set out to achieve. I think also, Chris, you might have asked there, you know, how we started, how the ideas came about. And for me, as soon as I was tasked with co-pilot enablement, my first thought was, you know, hey, this is an organizational change management problem. And how do we approach it like it was a change management problem? So

Walt Wear (03:35.23)
Of course, some of the absolute pillars in that area are really about communicating, making sure you’ve got a population that is familiar with what you’re trying to achieve, is comfortable with the changes, knows what to do with those changes. So I immediately set out to shape our adoption around a community-centric focus.

Another thing that we might touch on later at some point, but I’m I’m pretty passionate about is that we actually it’s been proven that we learn best from our coworkers. So there’s a ton of ways to learn things, but it’s come out in the evidence that we, out of coursework, self-paced learning, and a lot of other types of learning categories, we learn best from the people we know. so

They show us the things that are going to make us more productive, whether that’s in a traditional technology or something like Copilot. And I wanted to make sure our approach was based around that principle. So we wanted to engage the community, have members of the community sharing their experiences, their use cases, and mix that in with your basic educational aspects, showing someone where to click, how to use this feature. Really just

put ideas in the minds of our users about how they can be more productive with Copilot.

Chris Nichols (05:03.97)
So let’s get into some of the tools and and highlight some of your points that you just made. let’s start with office hours. Explain to the listeners how what how office hours are structured and how you run those, how often they are.

Walt Wear (05:15.335)
Yeah, absolutely. So office hours are our weekly community gathering point. So this is a a live call, a live forum where we welcome any and all questions, all and every level of experience to get together and learn something, share these experiences and ideas, and have a conversation. So

We originally structured this around leading with some educational content. So I think our first office hours, we looked at the new resources that we had just put out. We got into the standard basic copilot UI. We’ve gone through the different apps, how Copilot shows up in those apps, how to use those features for our productivity purposes.

After that section, we would have community members come in and speak to their own experiences, hold an open forum. We’ve since started to shift now that everyone has has access to a good base of materials for learning. We’ve started to lead with these community experiences. So I have had the privilege of having many speakers come on and join us from across functions in the bank.

And tell us how they’re using Copilot, what it’s doing for them, and really get into the specifics of their approach. The reason behind that is this idea of learning from our coworkers, this cross pollination effect. Even if you are a banker and you have someone who is in the technology space sharing an idea with you, they are this banker can kind of pull back a step from what.

this technical person is doing and find points of value and new ways to introduce those ideas into their personal workflows.

Chris Nichols (07:15.839)
And so one of the things that I love about it is you’re highlighting themes. So, whether it’s, you know, how to create a better PowerPoint, how to manage Excel, you know, and that’s kinda always gets me in that I that’s how I know, like, I need help on Excel, so I’m gonna tune in more. you record these and you put on AI Central with some good notes to them. so maybe talk about AI Central. What is that? ‘Cause I may miss a office hour, but I can go back and listen to a recording.

And then I’m always like, look at all these things. So talk about AI Central and the prompt library and what other resources we have there.

Walt Wear (07:51.752)
Yeah, so AI Central is our our main internet site for all things at the enterprise AI level. So when we put it together, one of the first things we did was set up that recording library so that people could revisit, like you said. Not everyone can make the meeting every week. maybe people just want the highlights, they just want to go read an article with the notes, they don’t want to watch the whole video. However, people prefer to consume that. We we wanted to make sure it was available to everyone.

To have access to for whatever their purposes are. With that, we also did introduce the idea of the prompt library. So before Microsoft introduced, they’ve introduced this new concept of an enterprise prompt library, which they have support for now. They’ve got personal prompt libr libraries, but we didn’t have that when we started this adoption, this adoption effort and journey. So we created a SharePoint list.

Obviously, that integrates very well with SharePoint sites, which AI Central is one of. And on there, we give people the ability to enter their effective prompts. We are able to categorize them by the application that prompt is used in, where it’s used, the job families that it’s going to be valuable to, how much time it saves. And we’re also introducing a social element there, including the ability to like these prompts.

You also get a like count. And naturally the prompts that people are most interested in can be sorted to the top so that people can find the most valuable prompts quickly. And we’ve gotten a lot of great feedback on that. It’s been really, really good.

Chris Nichols (09:37.847)
And what have you done to help us in terms of teaching copilot overall? So I think, you know, like the t template for PowerPoint example. So now I don’t have to load a template in. tell us about some of the behind the scenes work that you’ve done, that the team has done for Copilot.

Walt Wear (09:57.512)
Yeah, yeah. So I think what you mean there is the the preparation of materials, the making sure we’ve got assets that are backed at the organization level, that we aren’t introducing ideas that people will just run wild with. We want to give them the structure and the materials that would enable them to be successful or be most successful here. So, you know, like you mentioned with our our PowerPoint template.

We have branded assets that we want to ensure our customers are seeing. We want our teams to adhere to so that we’re able to put forth our best image at all times. And with that template, we realized through conversations with Microsoft through some personal testing that you need a template and you need great example slides to have copilot produce the best

Best outcome there, the best output. And fortunately, we already had this template. We had excellent slides, excellent slide examples. Copilot was easily able to build off of those examples for anyone who was building in PowerPoint. Now, beyond that, we do other materials preparation for our office hours. We put together on-theme use cases that are are usually totally make-believe, made up, but it helps illustrate.

a real world use case. So if you are speaking with a group of commercial bankers, it’s good to teach them how to use Copilot in Excel by using a an asset that they would be familiar with, something like a personal client tracker, things like that.

Chris Nichols (11:45.996)
and one of the other items, I think the third tool that I wanted to highlight was our Power BI overview, so now you can see what’s happening enterprise wide. Tell the listeners just how we monitor usage using that tool.

Walt Wear (12:06.157)
yeah, our copilot adoption slash usage metrics report. That has been a fantastic addition. It has made so much information more visible to our leaders. And what we’ve done there is connect on the back end to reports that you’re able to pull from act activity reports from Microsoft’s administrative backend. And through working that data, working with a really fantastic developer.

We’ve been able to generate reports that can break our usage, prompt number of prompts, number of users, activity dates, all down by our executive level, our department level, and down into management as well. And now everyone in those leadership groups is able to jump into their own metrics, their areas metrics, and make decisions about how they want to pursue adoption in those areas.

Chris Nichols (13:04.739)
Yeah, so an example of one of our our goals or visions of local market leadership, we push the some of those position those decisions down and that dashboard gives me as a manager kind of the tools to say, Hey, this person needs a copilot license, this person doesn’t. or, you know, not to say this is competition, but I am competitive and I do track Caleb, our our podcast producer, who has the most usage in our division. So I’m always kind of looking

How’s he using it? What can I be doing better? Et cetera. So it does get the competitive juices flowing and kind of helps managers decide like where they need to spend some resources on training. We don’t do enough in Outlook, for example. So we’re working on that. But that’s really helped me as a manager. So great work there. Walton, the in the past year that we’ve been at this, what are some of the lessons you’ve learned? So if a bank has listened to this and wants to kind of mimic or or use some of the ideas that you’re throwing out, how can you save them time and what

what you didn’t know a year ago that you know now through this experience.

Walt Wear (14:07.443)
Yeah, yeah, great question. I think that I would recommend first this idea of community, the idea, the concept that people learn best from one another. So I know I’m parroting myself from a minute ago, but those are really key elements. You want to give people a place to go where they can get the a source of truth about what’s happening with AI, what’s happening with Copilot. For us, that’s AI Central.

You want to give them a place to go where they can share ideas, where they can learn from one another. And you want to be intentional about the education you’re putting out there. You want everyone to have access to it. You want them to be able to ask questions. I think that copilot is such a unique technology because it’s not like other traditional technologies where it’s a a hey, this is how you click this button. This is how you run this process. Copilot, we have

infinite possibilities. And that’s why the idea sharing the community becomes so essential because if you don’t have that, it takes a great deal, enormous deal of creativity to come up with every new way you could possibly use Copilot in your world. If we left that up to each individual without this cross pollination, we wouldn’t be nearly as effective as we’ve become, as each user has become.

Outside of that, I would actually add one other thing that I think is a little bit, it’s gonna sound a little disconnected here, but it’s the idea of using copilot as a tool to use copilot. So it’s a little meta there, but it’s something that I actually regret not pushing sooner. And what I mean here is let’s say you’re working on a new task or you’re reflecting on the performance of a task you perform with copilot regularly.

It can actually be more effective if you come to copilot with the problem. So you ask copilot, how can you help me? rather than coming with a planned solution. Hey, I want this output. That’s saying I’ve got a solution that I want you to fulfill. Instead, if you go to copilot, when you’re dealing with new work, you can say, How can I best approach this with you? Let’s have a conversation about this idea. That is a real unlock.

Walt Wear (16:34.184)
For people.

Chris Nichols (16:36.533)
One of the I would s you know, now that you bring that up, that’s one of the things that I’ve evolved late in life too. When I first started it it was very static. Now I do have more conversations with co pilot. I often say, Hey, I wanna learn about how to use co pilot more in this area for this particular job function. We’re doing strategic strategic planning right now. So I asked us to kinda red team it, think through, play different roles, you know, play the part of you know, I have

hopefully this won’t get out in the public, but I have all my bosses kind of modeled in Copilot with their personas and personalities and I’ll say, Hey, play this part and you know, what what’s this individual gonna say so I can be better prepared and I often like put in like, how can I even do better? And it may reference something that you said on one of the calls. So I’m like gosh great, I’ll go back and listen to that on how to do it for a specific job function like financial forecasting or what have you. So

Walt Wear (17:32.02)
Chris, I hope that you let that out into public. I think that is such an incredibly powerful idea. You’re actually not the first person to mention something like this. I really like the level that you’ve taken it to, where you’re you’re asking Copilot to understand and you’re building out this persona. You’re understanding what your audience needs. Copilot now understands what your audience needs.

Chris Nichols (17:32.461)
That that is something fantastic.

Chris Nichols (17:37.353)
Ha ha ha.

Walt Wear (17:58.106)
That is going to make it more effective at communication. We all know that when we’re talking to our audiences, we need to know their needs to be able to meet those needs. And so I love what you’re doing there. I think that’s incredible. Like I said, you’re not the first person to mention a concept in this realm. and that’s actually something that I’ll take a second here to go on a tangent about. I think that oftentimes we will feel like

Is something we’re doing with co-pilot is is like cheating, or there’s like there could be like a stigma there. and I I think that’s such an important misunderstanding to break. If you can do it with copilot, and it’s an ethical thing to do, you’re not breaking your integrity down, do it. Any way that you can leverage copilot, leverage copilot.

It is like we talked about a minute ago, it’s a world of infinite possibility. You have no idea what value you’re gonna uncover and it could feel uncomfortable and unfamiliar at first. Lean into that until you get past it and see where it takes you. So I love that idea.

Chris Nichols (19:06.339)
So so stay with that theme. That’s actually interesting and a place I wasn’t planning to go, but let’s go there since you brought that up. There’s there was and I don’t know if it’s still going, but I th I think I still feel it kind of a backlash like, you used AI, right? And and now, you know, I talked to my team about, you know, you got you gotta do a human review on it and while you’re doing that human review, you know, train your co pilot to not sound so co co piloty, like take those dashes out, you know

on presentations it always puts that tagline at the bottom, which is fantastic. And I don’t I don’t kind of have a good clear vision about where to take that. What is your view on, you know, the difference between AI slop? So we all know we don’t want the AI slop there. We do the human review, but how much do you go from a human activity to a copilot activity and then turn it back into a human activity? You know, right? Like the like the dashes. Do I spend time

taking the M-dashes out so it doesn’t look like copilot. What’s your take on that?

Walt Wear (20:07.165)
Yeah, great question. There’s a there’s a big conversation broadly about AI slop. And I I know you’ve used and shared this concept before. And I think that, you know, that this idea of AI slop, I think it’s an appropriate term. It can sound a little bit harsh at times, but people are developing a skill set in detecting when your your communications feel inauthentic.

Because it’s riddled with these AI tra AI hallmarks. so what I think you’re asking there is, you know, how do I make sure that I am being authentic when I’m pr when I’m putting these things out into the world? And I think that’s an important thing to continue to ask ourselves, especially as bec as people become more literate with AI, as our our customers, our internal customers, we’ll be able to detect when

We’re we’re being we’re just doing too much through AI. I think the more undetectable AI is in your outputs, the the more human they’ll feel, which I think is the ultimate goal there. That said, I do think that AI is acceptable to use in in so many places. It just all comes back to making sure it’s authentic to you. So use Copilot to

Do more as yourself, through your voice, through your authenticity. And oftentimes we can use certain copilot tools to help us get there. Microsoft is plugging custom instructions into more areas now. We’ve got it in places like Excel, the regular copilot app. You can build out prompts that define who you are. one one exercise I actually did at a point was to feed it.

A lot of my inputs from Copilot office hours and ask it to understand my tone. So I create very extensive notes to make sure I’m hitting everything I want to hit in those meetings. And I gave Copilot that with the task of assessing who I am. How do I speak? How do I communicate to these groups? So that it could then help me in the future. Hopefully that covers what you were looking for, Chris.

Chris Nichols (22:27.565)
Yeah, absolutely. And I do want to underscore your point about removing the stigma because I’ve now evolved kind of through the looking glass and I’m now worried about the employees that aren’t using copilot. So if you’re if it looks a hundred percent human, I’m like, well, wait, didn’t you use copilot on this? Like, how much time do you spend in this document? Like, just use copilot. So I know when I assign work or when I ask for a goal or a mission, I’m expecting you to use copilot and or you know, and all our other tools to use. So I think it’s the opposite stigma that’s now evolving.

Walt Wear (22:57.555)
I think that’s fascinating and fantastic. I think I think that’s a good thing to be looking for.

Chris Nichols (22:57.872)
talk to us about you know, we’re we’re starting to transition into more agent and genic AI. Copilot Studio, you mentioned at start of this podcast. tell us how that’s going and it’s kind of a kind of a new area for us. So I feel like we’re almost starting over on our generative AI journey again, using A agents, but we’re trying to get adoption. What’s what’s the plan there and what are your thoughts around creating more use of AI copilot studio?

Walt Wear (23:33.096)
Yeah, yeah, absolutely. So in obviously when we’re talking about Studio, we’re talking about agentic AI. And one of the incredible things about Studio is that it can be a platform to interact with agents across other platforms, to monitor the activities of these agents. So there’s a ton of value in spending time for your organization to build copilot capabilities.

and by that I don’t mean solutions. I mean the capability to produce solutions using copilot studio. And I think in our case, what I’m most excited about is our embracing of citizen development. So we have a lot of smart people out there across the bank who, whether they’re bankers, whether they’re in risk, anywhere, they also have a set of technical skills that can be utilized in Copilot Studio.

So obviously there’s a lot of risk and a lot of oversight that we have to take into account. There’s potential costs. But the more, I believe that the more access we give to people to use and learn these advanced tools, the better our our solutions are going to become over time. I think I’m getting a little bit off of our direction now, but we are

We are embracing Copilot Studio as a core capability. Of course, with other vendors in consideration, we don’t ever want to, you know, pigeonhole ourselves. But really, again, we’re we’re we’re making sure that people have the environments, the security, the governance, the tools that are needed to enable citizen development. And that is what I am most excited about.

Chris Nichols (25:22.997)
It is a brave new world out there. So, you know, I think we’re all just learning how to best use these agents and you have the company has given me the tools. I look forward to more content from you about how to better use these agents. I’m just using, you know, starting to use loops and that’s something that kind of a I picked up in the last month or so to have my agent kind of redo its work and keep going without my intervention, without my feedback to it hits a goal, like producing a full report or like producing a full analysis.

you know, as we work with fintech partnerships, I always kind of feed information in and then ask the agent to create a plan for this particular partner. and I love that it has access to all my documents that you know we’re allowed to ha use. and so like it always brings up stuff that I forgot about. So always good. But I think the agents are are our new our our new paradigm to now i handle change management on and get the enterpr get every enter enter every employee

being a citizen developer to the extent that we can. So I know you got your work cut out for you there. All right. Well well, as we wrap up, as you look ahead, what are some of the things that you’re going to be working on? What are some of the things that you think banks should be working on overall, in terms of boost productivity. And also I want to also touch on how has this changed us in terms of FTE saves or better quality of work from your perspective, you know?

Walt Wear (26:27.853)
Yeah, looking forward to it.

Chris Nichols (26:51.245)
maybe tackle that question first, like what what are some of the things that you’ve seen as positive offshoots on this?

Walt Wear (26:57.691)
Yeah, absolutely. So when we look at the benefits from copilot, I think zoomed into copilot specifically, kind of letting the broader AI conversation stay out there, where we we could choose to go into a tangent about making sure that we’re working on improving our processes with AI, assessing the processes for feasibility of AI, that’s getting more in the agent realm. But when I look at that.

Personal productivity lens of Copilot. We have seen fantastic, incredible time savings. Now, I’ll qualify this with a. We assess this, we estimate this based on a conservative number of minutes saved per prompt. The I think that’s kind of the best you can do with Copilot right now without going into

Deep analysis of the actual prompts that are being used. That’s maybe possible. I would actually getting into that space, I would actually push back a little bit and say, is it really worth it? I think that this estimation of time saved, FTE saved, is enough. And I know that most recently, when we’re talking about 200,000 prompts, you can imagine multiplying that by three or five minutes and

Just starting to see how many hours are saved or how many FTEs are saved. On a monthly basis, I believe we are approaching over a hundred at this point per month. So saving ourselves the time of a hundred employees, which is just incredible. Yeah, yeah, I think that’s where I would leave that.

Chris Nichols (28:43.075)
Yeah, absolutely. And I and I will say there’s also the added benefit of giving everybody more confidence. I know my team has tackled tasks that it wasn’t really equipped to do, but with the aid of Copilot is now feels empowered to tackle those tasks. And so I think morale’s better overall and I think they’re more, you know, preparing for whether it’s a podcast with you or a client call, you know, we’re more apt to I think have the information at our fingertips and feel more confident about how we present ourselves. So I think

You know, you can’t underestimate that. But for sure I know I know communication’s gotten better, productivity in terms of just the way that every employee processes, you know, emails, produces presentations, produces memos, in have better quality policies. It goes on and on and on. But I think our work product is better as an organization thanks to, you know, AI and generative AI and genet AI. So thank you for your effort there.

Walt Wear (29:39.102)
I would absolutely and I do just want to say I completely agree with the quality aspect. There is a time-saved component. There is a quality aspect. But I really want to highlight you calling out this broadening of skill sets. I think that is one of the most incredible advantages that we get with Copilot is that we have comfortability in going outside of our comfort zone of trying new things.

I’ve heard a lot of people doing things that they probably would have never tried without Copilot. Things like writing Excel macros, advanced Excel macros to transform data in their Excel worksheets. I know someone that took that a step further and did something with Power Automate. They used Copilot to learn Power Automate so that they could perform now this transformation automatically. They’ve got the script that runs in Power Automate.

They’ve got a rule and outlook that causes that report to hit the SharePoint site, triggers the power automate flow. And they did this not even for themselves, but for a coworker. And they saved, I think they estimated on average, something like two hours a day or s or something very dramatic like that. None of this, it would have taken months for this person to learn that skill set and be able to accomplish those outputs and outcomes without copilot. So

Thank you for mentioning the broadening of skill sets. That’s a huge one. And the return on time as well.

Chris Nichols (31:08.43)
Yeah.

Success abounds. okay, what’s ahead? Our exit question. as you look ahead for next year and as you do next year’s strategic planning, what what do you want to get accomplished for us?

Walt Wear (31:22.193)
Yeah, big question there. So I’ve recently accepted a new role as our AI enablement manager. And the purpose here is to take what we’ve done so far and go a step further. I will be more involved in our intake process, everything from business case development all the way to benefits realization management. And that benefit realization management is such a key aspect.

As we move into these AI solutions and outside of this personal productivity space, into this department level change or enterprise level change, business process level change. These are things where we should have measurable KPIs already in place. And we can start to assess how much of an impact these new solutions are having. That is what I’m most excited for, is

Doing more in that space of helping people articulate what they want to accomplish with AI, what they think is going to be beneficial for their business area, putting measurements around that, getting that through to deployments, and then being able to assess how well we achieved our target. So we can share with our internal stakeholders as well as our shareholders how much value we’ve been able to create through our AI adoption.

option efforts.

Chris Nichols (32:49.167)
Yeah, and for those listening, listeners that have, you know, learned our strategic process, our shark type tank like system, what Walt just described is is core to that and and what we say is, you know, capital flows where capital’s loved. And, you know, every year coming up in September, October, you know, as a manager, we have a chance to say, you know, how we’ve done and to to Walt what Walt alluded to, our accountability towards benefit realization on the projects that we said.

That we’re going to do last year and the year before that. And so you’re held accountable. And that’s one of the things I love about it, because you know, you can see your shortcomings and everyone can see your shortcomings or your successes. And then they can help make the decision about, you know, whether you want you deserve more capital and how you’re going to grow your business. and you know, Walt’s what Walt just described is key, particularly now. We’re all scrambling to try to figure out how best to use AI. And, you know, the track records are gonna vary, but we’re all gonna learn from each other to Walt’s.

point about a community and and then you know get it right hopefully have be better allocators of capital at the local level to drive our you know remarkable experiences for our clients and our our our relationships for with our employees. Walt thank you so much for being on thank you for all that you do for us. Thank you for giving us the tips. Hopefully other bankers you know this will generate questions and other bankers can learn from you. We’ll be feet you know want to feature you again maybe check back with you.

next year. We’ll also probably do some a couple companion pieces on, you know, how banks can better leverage Copilot and how banks can better get their plan out to have change management within the organization to use it. So thank you much and we look forward to hearing about your journey next year.

Walt Wear (34:32.655)
Thank you so much, Chris, for having me. I just wanna say it’s been an honor and I can’t wait to see how we’re able to help other banks in our network. So thank you so much for having me.