Why Finance Needs to Find the ROI on AI — and 3 Ways to Get There Fast

If you struggle to find the ROI on AI spend, you’re not alone.

It’s a problem for most finance pros — and it’s only going to get more difficult as a growing number of organizations are demanding to see a positive link between the two.

Eight-seven percent of finance leaders say they must tie AI spend to business outcomes within the year, according to a CloudZero survey.

Here’s the catch: Just 22% can actually do that today.

The AI Paradox for Finance

“We call this trap the AI paradox. Like it or not, companies must spend more on AI to stay competitive, but the spend itself has become ungovernable,” the CloudZero researchers noted. “Finance can see the activity, tokens consumed, models called, the monthly bill, but can’t connect it to the customers, products and transactions that money was intended to serve.”

At this point, 60% of finance pros admit that they’re spending more on AI than they can justify. Two-thirds of boards already say they condition further AI funding on proof of return. Another 22% will approve nothing new until ROI in AI shows.

Whether it’s the CEO or a board asking the questions about AI spend, the answers need to be serious and steeped in quantifiable data.

Naturally, the ROI is equally important to finance pros: Nearly two-thirds of them say that if they can tie spend to the outcomes, they would fundamentally change how they invest.

What’s Happening When There’s No Connection

When the outcomes can’t be proven, the money stops.

According to the research, 75% of finance leaders who can’t show the ROI of their AI have held back on investment, and 35% have killed or paused an initiative. For those who can prove it, those numbers drop to 38% and 11%.

Of course, without full insight on how well the AI initiatives have or will go, holding back or killing the project might be the right decision. On the flip side, without ROI insight, organizations might be missing out on the right initiative.

Connecting the ROI on AI

“The companies that solve the AI ROI equation are funding what works while everyone else stalls,” the researchers said.

Here are five tips to move closer to showing the ROI on AI investments and use.

1. Close the gap. Where and when possible, gather spend data on day one. When spend reaches finance within a day, it almost “behaves like a different category than data that arrives several days later or with the bill,” the CloudZero researchers said. If you’re like 80%+ of other finance pros who have to prove ROI on AI within a year, start daily data now.
2. Give everyone the same view. The people who are spending on AI and the people accountable for it aren’t usually the same. For instance, many front-line workers use AI, but they don’t understand the cost of tokens consumed. Give everyone a shared, real-time picture of how use is connected to spend, and you might be able to lessen the gap without uprooting protocols.
3. Look at one painful workflow. In Finance, that might be close support, variance analysis or AP coding. If you want data from all businesses, ask the unit leaders to do the same. Then baseline it before changing anything. Track current cycle time, error rate, escalation volume and hours spent so every gain (or loss) made with AI has a clean comparison point.
4. Be multifaceted when you measure ROI. Stick to three lenses: efficiency, quality, and speed. For most teams, that probably will be hours saved, less rework, and faster close or forecast cycles.
5. Measure spend against outcomes. Try to get a granular look — cost per the thing each business runs on. For instance, per customer, per feature, per dollar of revenue. With that perspective, you can pour more into what’s paying off and pull back from what isn’t. “That’s what turns a spend report into a decision and an action,” the CloudZero researchers said.