How businesses can understand the ROI of AI

This is the print edition of Forbes' CFO newsletter, providing the latest news for budget-focused chief financial officers and other leaders. Register here Delivered to your inbox every Tuesday.

---

As we approach the latter half of 2024, many CFOs share a common top strategic imperative. Use AI and other automation technologies to perform tasks previously performed by humansA second-quarter CFO survey conducted by the Federal Reserve Bank of Richmond, the Federal Reserve Bank of Atlanta and Duke University's Fuqua School of Business found that nearly six in 10 companies have already implemented software, equipment or technology to perform these tasks, and 53.6% of companies want to add to their tech stack to further automate employee tasks in the next year.

However, the point to note is For now, most finance departments don't want to replace those people.More than 87% of companies that implemented AI in the past 12 months did so to enhance business processes, roughly the same percentage of companies that plan to use AI for this function in the next 12 months. Improving the quality of deliverables is also important: 57.7% of companies deployed AI for this reason in the past year and 58.1% plan to reap the benefits in the next year. Fewer than half of companies hoped to reduce labor costs through AI in the past year, but 54.7% cite it as a goal for next year.

Daniel Weitz, director of research at the Federal Reserve Bank of Atlanta, wrote: The current economic situation may push CFOs further towards AI automationAs inflation rose in 2021 and 2022, so did input costs and employee wages, which led to higher unit costs. Though the pressure has eased recently, price pressures remain quite high. Survey results show that companies that decided to automate tasks last year also saw faster price growth then. And those same companies expect price growth to slow this year, likely because automation and increased accuracy from machine-based calculations will lead companies to spend less.

“Time will tell whether the polarization of price growth expectations across the automation landscape is borne out and whether companies that automate return to 'normal' price growth more quickly than those that do not automate,” Weitz wrote.

Most companies AI will contribute to revenue, but it's hard to know how much, and how long it will take.I spoke with David Obrand, CEO of AI-powered sales engagement platform Salesloft, about what to consider when making that calculation; excerpts of our conversation appear later in this newsletter.

Stock Market News

The stock market saw a series of big gains and modest declines last week, thanks in large part to Nvidia. The S&P 500 hit 5,500 for the first time in history last week, with many tech stocks surging thanks to a surge in AI chipmaker Nvidia. Temporarily The world's most valuable companies The company took the top spot last Wednesday. It lost its spot to Microsoft on Thursday and dropped to third place behind Apple on Friday. After a surge last week, Nvidia's stock quickly lost steam, falling below $120 a share this week.

Nvidia is the latest beneficiary of the stock market's AI boomThe promise of AI technology is propelling many big technology companies to new heights. Apple's upcoming announcement of AI integration into its smartphones, tablets and computers, and Microsoft's positive analyst reports helped catalyze a stock rally last week that propelled NVIDIA into the stratosphere. But NVIDIA also announced big news last week, partnering with Hewlett Packard Enterprise to unveil a co-developed portfolio of AI solutions and integrations to help companies rapidly adopt generative AI.

While other companies are building AI platforms, Nvidia is developing the chips needed to run these platforms and the data.CNBC estimates that Nvidia controls 70% to 95% of the AI ​​chip market. Microsoft accounts for 15% of Nvidia's revenue, according to Bloomberg supply chain data. Sequoia Capital estimated in March that companies were spending $50 billion on Nvidia chips to train law students. The Wall Street Journal report.

This does not include NVIDIA's upcoming Blackwell platform will feature its most powerful chips yet.At the time of its announcement in March, Nvidia CEO Jensen Huang said it had already secured commitments to use the new chips from Amazon Web Services, Dell Technologies, Google, Meta, Microsoft, OpenAI, Oracle, Tesla and xAI. So Nvidia's valuation and revenues should continue to climb even if development of its AI platform stalls. Nvidia's annual shareholder meeting is on Wednesday, which could bring more market-moving announcements.

Legal issues

The U.S. Supreme Court last week Individuals who own shares in offshore companies may be subject to tax on their earnings, even if the money remains within the company. The main issue was whether Parliament had the power to enact the law, which was challenged by a married couple who held shares in an Indian company, who argued that the law taxed their personal assets rather than their actual income.

The ruling in favor of the couple – that unrealized income would not be taxed – could have had a major ripple effect on general taxation in the United States.The federal government will likely lose trillions of dollars. Forbes Senior writer Kelly Phillips Erb spoke with several members of the corporate tax community about the decision, and while many applauded the decision simply because it kept the current system in place, others expressed concern that the Supreme Court was interested in hearing a case on the issue in the first place, and feel it could lead to further litigation that could result in changes to the tax system.

Policies + Regulations

IRS (Internal Revenue Service) Plan to deny companies tens of thousands of improper high-risk employee retention credit claims In the aftermath of the COVID-19 pandemic. Forbes Senior writer Kelly Phillips Erb writes that the program was intended to help businesses survive the early stages of the pandemic, when many businesses were forced to temporarily close. Since its inception, businesses have filed 3.6 million claims against the program, costing them more than $232 billion. The IRS says it has investigated 1 million claims worth $86 billion so far and identified 10% to 20% as high-risk with “clear signs of being false claims.” Another 60% to 70% of claims show “unacceptable levels of risk,” which the IRS plans to analyze more deeply. The IRS says that these claims are complex and will take time to investigate, but that businesses who feel they may have filed suspicious claims can apply to have them withdrawn.

Off the books

Salesloft CEO David Obland explains how to calculate the ROI of an AI platform

If you're planning to add an AI platform to your enterprise, the ROI can be hard to quantify. We spoke with David Obrand, CEO of AI-powered sales engagement company Salesloft, about some things to consider to get some realistic numbers. This conversation has been edited for length, clarity and continuity.

Generative AI platforms are being deployed to help companies help people do their jobs more efficiently. But how do you quantify that? If someone comes to you and says, “I want to do this, but I need numbers,” how do you quantify it?

O'Brand: We're looking at the KPIs that matter most to go-to-market organizations: cycle time to actually close a deal, deal size, deal win rate, everything. These are all things our customers want, especially in today's tougher macroeconomic environment, where interest rates are rising and companies are cutting back on spending and cautious about allocating capital. The bar to revenue growth is much higher for all kinds of organizations than it was in recent years. Now, everyone is looking at “how can we be more effective with less?” The answer is, “how can we identify more opportunities by incorporating digital signals?” Today's buyers are more digitally empowered than ever before. [been] in front.

When companies consider where they can create value with AI, is there anything they tend to overlook?

It depends on the use case. To train an AI model, you need to ingest a lot of data. So you need to give a lot of thought to where that data is coming from and what the data privacy laws are related to it. For certain use cases, you need to be mindful of the ethics around the use of AI. You also need to train your users on how to use AI appropriately. I think that's one of the big areas we're starting to see today.

There are many examples of AI being rapidly adopted without proper training from users. The AI ​​itself is not bound to the truth; it is simply trained. [data]You can also apply generative AI to listen to conversations and write follow-up notes. In fact, we generate three templates for every call. Salespeople can review these templates, validate them, select the parts they need, and run them together. What if we just generated an email summary and that was one version and nobody saw it? We just sent it. If the AI ​​hallucinated or simply missed a core component, the person who received my email would think, “Oh my goodness, this person didn't pay attention to what I said at all and didn't understand it at all.” And that's it. So we need to make sure that we are training the users of the technology in whatever form. I think data privacy on the one hand and user training on the other are areas where people can take shortcuts in their enthusiasm to their detriment.

What arguments should CFOs expect to hear from those trying to prove that AI platforms are worth the investment?

Our world is going to emphasize the ability to generate revenue with much greater consistency and repeatability than in the past, and when you break that down into silos, you look at it and say, “We can reduce our marketing spend and generate more pipeline than we did before.”

The second area is generating more revenue with fewer resources. AI doesn't replace humans, but it enables them to do their jobs better. If everyone is performing well, you don't need as many people. You can think of it as revenue per sales rep, for example.

If you're in a competitive market, you can look at it as win rate or market share. Can we capture a larger market share? And because you're able to bring in first-party data that can be extremely helpful in helping your customers cross-sell and upsell to their customers, you can look at it as: “How much improvement in wallet share?”

Facts + Comments

According to a new report from Citigroup, the banking industry Will be heavily influenced by generative AI.

54%: Time spent on banking tasks that can be automated

$170 billion: The study estimates how much AI could contribute to the banking sector's profits by 2028, which would amount to a 9% increase.

“Revolutionize the banking industry and improve profitability” Citi's Chief Technology Officer, David Griffiths, said this about what AI can do:

Strategy + Advice

As The CFO's job is changing, and so are career paths How do you progress from a finance employee to the top position? Here are some tips from experienced CFOs to help you find your way there.

It's hard to be productive Here are some tips to help you achieve more at work:

video

quiz

Danish pharmaceutical company Novo Nordisk this week It plans to build a $4.1 billion manufacturing facility in the United States. It is a manufacturing site for the weight loss drugs Ozempic and Wegoby. In what state is this facility located?

A. North Carolina

B. Texas

C. Kansas

D. New York

Check here if your answer is correct.