The CFO AI Reality Check: How Finance Teams Can Make AI Work
Most finance teams are already spending money on AI. Very few are seeing anything come back from it.
That gap was the focus of The CFO AI Reality Check, a webinar we recently co-hosted with our partner Anaplan. Mavens Cloud Chief Technology Officer, Nils Beecken joined Adrian Cadman, Senior Director, Finance Applications at Anaplan, for an honest conversation about why so many AI projects in finance fall short, and what the successful ones do differently.
Here are four ideas from the session that stood out to us.
1. It's a measurement problem, not an adoption problem:
Adrian opened with research from Gartner. Around 59% of finance organisations are actively investing in AI. Yet 84% of them aren't seeing the benefits they expected. Nils summed it up simply: "Adoption is high, realised benefit is low. I don't see it as an adoption problem. I see it as a measurement problem."
Many businesses never set up a way to track what AI is giving back. And often the AI is running on data, the team doesn't fully trust. If you aren't confident in what goes in, you can't be confident in what comes out.
2. Finance needs answers that don't change:
AI tools like ChatGPT are great with words but ask one the same question twice and you'll get two slightly different answers. In finance, one plus one has to equal two every single time. Nils's view is that GenAI is not a calculator and can’t be. Build it on a trusted planning platform where the numbers are solid and let GenAI help people find and understand them.
In the webinar, Nils also walks through the four stages finance teams move through on the way to GenAI that can act on their behalf and explains why skipping any of them is where projects tend to fall over.
3. Don't wait for perfect data:
One attendee asked whether messy data means a full clean-up has to come first. Nils didn't hold back: "If you do a full clean-up first, you will fail."
Instead, pick one process your team knows well, clean the data for just that process, and choose something where you can check the results quickly. In his words: "Pick one, clean the data, prove it works and then move forward." One early win builds more belief across a business than any big promise.
4. Treat it as a people project:
Gartner suggests around 60% of AI projects in finance either stall or fail. Nils shared three simple ways to avoid becoming part of that number, and none of them are about technology. He also spoke about the very real worry many people have that GenAI will take their jobs, and how training and honest conversation can turn that fear into confidence.
Our take:
GenAI is moving fast, with new models arriving almost every month. That can feel overwhelming, but it's also a real opportunity. At Mavens Cloud, we believe the finance teams that come ahead won't be the ones who move fastest. They'll be the ones who start on solid ground, prove value early, and bring their people with them.
There's no need to be afraid of GenAI. There's every reason to be thoughtful about how you begin.
Watch the full conversation:
Hear the full discussion between Nils and Adrian, including the four-stage roadmap, Nils's three rules for AI projects that deliver, and the audience's Q&A.
Watch The CFO AI Reality Check on demand
Want to talk about where your finance team is on its AI journey? Get in touch with the Mavens Cloud team by emailing enquire@mavenscloud.com.