CPA Firms and CFOs Share AI Implementation Strategies in Finance Panel

Heat. Heat. Heat. Heat. Thank you everyone for joining us today. Uh good morning, good afternoon, good evening wherever you are today and welcome to today's discussion. Uh my name is Chris Gainer. I'm the chief marketing officer here at Magentic and I'll be moderating the discussion uh as we go. Uh today's session is meant to be designed sorry it is designed to be practical. Um there'll be some data points and some slideware up front but very minimal just to frame the session and we want the session to be as interactive as possible. So please ask questions, share comments as we go. Uh and so on that note to test the interactivity. Um we've got a little poll to kick us off. Uh this is to understand where everyone is in the room around on their where they are on their AI journey. Uh so uh over on the poll section you'll see some questions. Uh please cast your vote. And as that's happening, uh, I'm joined by some excellent panelists today. And so, uh, whilst you're casting your votes, I'll throw it over to the panelists to get them to introduce themselves. Uh, John, can we start with yourself?

Yeah. Uh, happy to be here. I run a CPA firm in um, Utah, but we have clients and team members all throughout the United States, about 290 team members. Been around for about 15 years. Uh we do tax bookkeeping uh payroll, all the stuff you would expect an accounting firm to do.

Perfect. Uh thank you, John. Inside tax. Uh Mark, over to yourself.

Good morning. Uh my name is Mark Pritchard. I'm the CFO of a business called Har Mobility. We are based in Sarasota, Florida. We're a private equitybacked manufacturer of stairlifts and mobility equipment. We have um a site here in Florida. We have a site in Missouri. We also acquired a business in Northern Ireland uh two years ago. So, we're an international rapidly growing business. And um we tend to try and use AI as much as possible. I'm no AI expert, but I'm certainly not afraid to try things and try them again until we get it working.

Perfect. Um, expecting some great conversation then, Mark. And last but not least, David, over to you.

Thank you, Chris. Uh, my name is David Brain. I'm one of the co-founders and CEOs of Magentic, and we bring uh, we support our clients with implementing AI and outsourcing in order to help them deliver more effective business processes.

Perfect. Uh, Mark, John, David, thank you very much for joining us today. Uh, I think we've given enough time to the votes now. I can see. Yep, we've got a bunch of people who have cast their vote. If I click this, does it work? So, uh, not surprisingly, um, we've got majority of the room researching, experimenting. Um, this is pretty, uh, consistent with what we're hearing in the market when we talk to people. Uh, interestingly, is the not started portion. I would love to pick up the phone and talk to everyone in the room here and understand kind of where they're at, uh, with that thought process. Evaluating quite low and then running pilots is also interesting. It's moved up quite a bit from last time we ran this study. So, it's interesting to see that people are starting to move from experimentation into pilots and running in production also quite high. Um, super helpful for framing the room. Uh, thank you everyone for for filling that in. Uh, I will give one slide and a 30 second plug on Magentic and then I'll move into setting the scene for today. Uh, but quick plug for Magentic. As I said, Magen Magentic helps organizations go beyond experimentation as we saw a large portion of the rumor there. uh moving from a few Claude and Chatbt and Gemini licenses to running fully orchestrated AI plus people operations that deliver tangible value. Uh building on decades of experience delivering AI and automation programs for leading companies, Magentic brings together world-class talent uh with four deployed engineers to effectively identify use cases, design agentic operating models, establish credible business cases and critically get AI deployed into operations. Um really importantly here is that uh we are not tied to a single tool or vendor. So we work across a range of technologies uh always looking for the right tool for the right job at the right time. It's the end of the plug. Uh hopefully that was painless but if you want to find out more about who we are, what we do or some of our work, please reach out after the session. Right. Uh let's get on to the framing and we'll get into the discussion. Uh so Deote recently published uh some research that said 87% of CFOs say AI will be extremely or very important to their finance operations in 2026. Not interesting, not worth watching, but extremely or very important. Uh and more 54% say integrating AI agents into finance will be one of their top transformation priorities this year. Um so the intent to adopt AI is very clear. Uh Steve Deluchcci at Deote CFO program has said um leaders are looking into AI not just to improve efficiency within finance but to help their organizations respond faster to changing market conditions and customer dynamics. Speed is critical. The question we hear most ourselves is isn't so much where should we invest in AI but where do we begin? And so um with that I'm going to dive into our second poll straight away which is around use cases. And it's just to understand everyone in the room as you're evaluating or experimenting with AI what is the top so not there might be a few competing here but pick just one what is the top use case that you are exploring within your organization um and at this point I wish I had some music us through but I will fill it with words as I ramble slowly slowly to the close of the poll. Uh we're getting a few people fill in here. I give everyone a couple of seconds to submit here. Couple more seconds. Uh Trey has it said, "Preaching to the choir here. How do we do this?" We will hear more today, Trey. Um and then if you've got something specific, obviously give us a shout afterwards. All right, we've got a few submissions here. About yeah, more than 50% have answered here. So, let's have a look. Um automating simple tasks looks like the winner here. And it's pretty consistent with what we saw last time we ran this poll. Analyzing, extracting data and documents, document processing, data proc processing really critical summarization critical um as we go down complex decisions and other I'm always intrigued by that other box to find out what actually is happening there. But when we ran it last time, just to show uh a little comparison to last time, pretty consistent and we did this probably about a year ago. And so whilst the the zeitgeist have you believe that people are much further along um you are not alone where you are and at this point so the data is telling a story which is uh the top use cases are typically not the most glamorous automation of tasks uh document processing summarization fundamentally it's all of the elements that enable the operation um and so again this is sorry I've failed to click there this is the use cases we saw last time as well very consistent with what we saw today And so while the aspiration to adopt AI is quite high, the uplift as you'll see from 23 to 24 here dramatically went up yearon year in terms of percent adoption. From 25 to 26, the adoption has fundamentally slowed. Gartner believes uh that people who are further along in their journey, so past the experimentation phase into um rolling outlets or even further are 23% more optimistic once they've gone through the hurdles and they've understood what it takes to get it into production than people who are at the start of their journey which are around 7% um uh feeling confident. So clearly getting started is the hardest part and critically it requires us to look at some hard truths. So this is the last slide before we dive into dive into the discussion but getting started is hard. First 91% of organizations report low to moderate impact initially in the early stages. Uh second 48% of CFOs site resistance um resistance in terms of uncertainty and doubt as their biggest challenge. So not technology, not budget, but people. Uh and third, 70% of AI projects simply fail to meet expectations. And typically it's due to data quality, talent gaps, overestimation of what the tech can do on its own without addressing the underlying processes and governance. And so food for thought here. I'm going to uh shift into the discussion here. Um so you can close down the slide portion and and move the the panelists into into frame if you wish. But um why don't we start with a real simple question for the panel which is I've used AI about 30 times in this discussion so far. It means a lot to a lot of uh to a lot of different people. Um how do we define AI today in the context of finance andor accounting? Uh John would you mind kicking us off?

Yeah. Um I mean just like the poll revealed for us AI is how do we automate some of the simplest tasks that can be a bottleneck that don't require really any brain power or the need for a human to be involved with a client experience. Um, but it's interesting. I was also I heard Steve Waznjak uh just the other day in a a commencement address say AI stands for actual intelligence and as humans we have AI. Um I I think there's going to have to be that balance between what can we get the AI as in the software and the robots and all the stuff behind the scenes. How do we pair that with still the uniqueness of us being humans? I think that's going to be an interesting conversation when we have and and going into the future and probably in this meeting as well.

Awesome. Actual intelligence. Love it. Mark, how are how are you defining it or at least explaining it?

Um, I'm just encouraging my team to embrace AI and try and use it. And a good example of how I want my team to use it. Um, I've got my 10,000 hours of Excel experience. Uh, I also have junior people who will never need to get