Why do most AI projects fail? Because enterprises haven't calculated a proper account.

Companies that successfully apply AI are not necessarily those with the most advanced models or the largest datasets, but those that can bring together various expertise and make the wisest decisions. With their ability to evaluate value, ensure accountability, and provide an objective perspective, finance teams play an indispensable role in ensuring that AI investments yield real returns. By involving finance teams early and frequently, companies can transform AI from an exciting possibility into a reliable growth engine that can both drive revenue growth and achieve operational excellence.

Enterprises are adopting AI at an almost unprecedented pace, but there is still a significant gap between enthusiasm and actual results. In a recent survey of 750 executives (150 from the technology industry and 600 from other industries), the vast majority of respondents (65%) said they believed they had an in - depth understanding of AI and its benefits, and 18% considered themselves to have a cutting - edge understanding. However, only 6% claimed to have the ability to extract value from these technologies and have a substantial impact on the company's profit and loss.

Our research and experience show that the performance of AI depends more on how, when, and where decision - makers apply AI rather than the technology itself. The success of AI requires close cooperation between business and technology leaders, rather than being dominated by one side alone. However, we hope to delve deeper, understand this decision - making process, and find out what exactly causes the difference between success and mediocrity or failure.

Therefore, we asked the business and technology leaders participating in the survey to rank the key areas of their company's AI investment according to Michael Porter's value chain model. They were asked to select the top three priorities from a list that included five main areas (procurement and supply chain; production and operations; fulfillment and distribution; marketing, pricing, and sales; and customer insights, service, and experience) and four supporting activities (human resources and talent management; finance; product, process, and technology development; and risk and compliance).

The research found that the finance department plays a crucial role in the collaboration between profit - and - loss accountability and functional department leaders, and this collaboration is the key to AI success. First, the data shows the key directions of companies' AI investment:

Companies focus their investments on

areas that can create the greatest value

Among all the surveyed companies, the top area for AI investment is customer insights, service, and experience, with 53% of the companies listing it as a priority. The second - largest investment direction is operations and production. The third is product, process, and technology development.

These priorities vary by industry. Capital - intensive industries such as aerospace, energy, and manufacturing rank operations first, followed closely by product and process development. In consumer goods and retail B2C enterprises, marketing and customer - related activities are far ahead, followed by the supply chain.

These priorities make intuitive sense. For capital - intensive companies, operations and processes are where they spend the most, invest the most capital, and create the greatest value; for B2C companies, the key to value creation lies in finding and serving customers and ensuring sufficient inventory on the shelves.

Leading growth companies prioritize the finance area

When we focus on high - performing companies measured by revenue growth (approximately the top 16% of the sample), a remarkable change occurs. For growth leaders in all industries, the priority of customer - related activities increases by 4 percentage points, from 53% to 57%; but financial activities jump to second place, replacing the priority of operations and product and process development. 53% of growth leaders choose finance as a priority area, while the proportion among all companies is only 37%.

The same situation occurs in industry groupings. In capital - intensive industries, the role of finance increases by 10 percentage points, from 33% to 43%. In the B2C category, this proportion rises from 24% to 36%. In the healthcare industry, the importance of finance increases significantly, jumping from 36% among all companies to 64% among growth leaders.

The finance department helps the company

make better AI decisions in three ways

Obviously, when the finance team participates in AI decision - making together with business and technology leaders, a kind of "magic" occurs. Here are the benefits that this collaboration brings to the company:

1. Select the right projects

Among the numerous possible AI application scenarios, choosing the right use cases is often a challenge. Our cooperation with clients shows that the involvement of finance makes the discussion between the technology team (what AI can do) and the business team (what needs to be done) more efficient. Finance provides a crucial third dimension: what is the most worthwhile to do?

The CFO's team is usually good at quantifying the return on investment (ROI) of projects. They can measure the impact of projects on costs, revenues, cash flows, etc., and calculate how these factors affect enterprise value. By using tools such as scenario analysis, the finance team can also help business and technology colleagues model and compare the effects of different project portfolios, while showing how various risks affect value creation. When the leadership faces multiple seemingly attractive projects, finance can provide an objective "touchstone". Some seemingly insignificant projects often bring the greatest return on investment. For example, for a company that provides preventive maintenance services for truck and other vehicle fleet operators, it turns out that an ordinary - looking project that uses AI to improve the quality assurance of software code is ultimately more valuable than many customer - facing projects.

In other cases, the finance department, with its unique advantages, is more likely to find opportunities that are most suitable for using AI for data processing. We once cooperated with a consumer goods company to improve its profitability and deeply realized this in the process. The company proposed dozens of AI project suggestions, covering aspects from optimizing marketing spending to reducing customer churn. Among these feasible suggestions, the finance department played a key role in ensuring that the focus was on ideas that could have the most tangible impact. Finally, the most important project was to use AI to optimize promotions and pricing simultaneously, that is, to determine whether promotional products need discounts and the extent of the discounts. This is a comprehensive project that combines advanced financial mathematics, the practical experience of the business team, and the technical creativity of AI experts. Eventually, the selected project increased the profit margin by 10%.

2. Solve the right problems

AI can confidently provide correct answers to the wrong questions, which makes organizations wonder why their investments are not achieving the expected returns. For example, assume that a company is facing the problem of customer churn, which leads to a decline in sales and an increase in the cost of acquiring new customers. There may be many reasons for customer churn: Are sales representatives more proactive with existing customers? Has the customer success team ignored some customers or missed warning signs? Is there a problem with customer segmentation? (That is, were the wrong customers introduced from the beginning?) Is there a problem with pricing? Is the customer service center inefficient or ineffective? AI can help solve these problems, but which problem is the most critical?

To conduct this root - cause analysis, AI experts, operations experts, and personnel with financial analysis skills (i.e., experts in the finance department) need to work closely together to find the root cause of the problem and develop targeted solutions.

"We often communicate with CFOs because they usually truly understand the problems that need to be solved," Chris Satchell, managing director of the technology and digital department at Clayton, Dubilier & Rice private equity firm, told us. He gave an example: A building materials manufacturer they invested in had been having difficulty effectively allocating funds and planning resources according to customer needs. Eventually, it was the finance team that found the root cause of the problem: the company was using a demand forecasting model that could only predict three months ahead with low accuracy. Due to inaccurate forecasting, the procurement, operations, distribution, and even sales departments were unable to make precise decisions. And forecasting is exactly where AI excels. They introduced a new AI model that can predict demand for the next 18 months, reducing the error rate by 50% and providing refined factory - level forecasting results. Satchell said that this provides the company with a powerful tool to optimize capital allocation and improve profitability.

The finance team has another advantage in identifying potential problems: they dare to challenge the "sacred cows" (things that are considered untouchable), face problems directly, and will not be as timid as other departments. They will guide everyone to have discussions based on facts, not being restricted by inherent ideas and biases. Therefore, they can also build a bridge between the project team and the executive team. If a project is worth funding, the CFO can usually find funds within the company, even if the budgets of the IT department or the business team are tight.

3. Ensure accountability and continuous value creation

Finally, the finance department can ensure accountability, which brings three major benefits to the company:

First, this is particularly important when selecting and managing suppliers. The AI market is highly competitive. The finance team can ensure that negotiations with suppliers always focus on the company's standards and value - creation goals, rather than letting suppliers set the measurement standards for you.

Second, the finance department helps protect existing gains and prevent their loss. Whether it is an AI project or other types of projects, there are many potential factors that may lead to value loss, such as project scope expansion and time overruns. If the gains (such as cost reduction and revenue increase) are not recorded, captured, and accounted for, value may also be lost. Especially when faced with dazzling new technologies, the team may be tempted to constantly test these gains, but this may not be the right business decision.

Third, the finance department can effectively manage the budget. This year's AI projects will soon enter next year's planning cycle. The involvement of the finance department can bring the following benefits: ensuring that this year's results become part of next year's budget; providing continuous funding for new projects; integrating AI into business and technology planning to transform short - term gains into long - term capabilities. In addition, the finance department usually has a broader perspective of the enterprise. The finance team involved can identify cross - functional or cross - business AI opportunities that are difficult to see within organizational silos.

How to achieve this goal

The ideal AI collaboration model includes three key conversations, combining what is needed from a business perspective (what needs to be done), the reasonable excitement of the technology team (what AI can do), and the balanced view of the finance team (what is the most worthwhile to do).

Compared with the technology itself, the team is always more important, and the most critical factor is the quality of the conversation. Only through high - quality conversations can AI solutions that can produce substantial results be identified and selected. To ensure that these conversations can proceed smoothly, companies can take the following steps.

First, define the criteria for success, whether it is to improve operational efficiency, enhance operational effectiveness, or drive revenue growth. These strategic guidelines will help keep the collaboration focused. Second, recruit a finance team member with the right mindset from within the organization. As Oscar Wilde said, a cynic is "a man who knows the price of everything and the value of nothing", and this is especially true for some financial personnel. However, if the finance team has the right mindset (and incentive mechanism), they will be able to provide insights that operations and AI experts cannot obtain, including when to decisively abandon a poorly performing project (breaking through the sunk - cost trap) and when to increase investment when the data starts to show results.

Finally, involve the finance team in its own AI transformation. When AI is applied to the finance function, the finance department can become a more effective strategic partner for the enterprise - not just a "historian" who views the business through the rear - view mirror, but a partner who can monitor business activities in real - time, enhance forecasting capabilities, improve the speed of insight, and reduce risks. As more and more companies automate routine financial functions, the application of machine learning and AI has become more and more common. Some companies have liberated talents through these newly gained efficiencies and opened up more creative and strategic career paths, such as collaborating with technology and business experts to widely apply AI across the enterprise.

Ultimately, companies that successfully apply AI are not necessarily those with the most advanced models or the largest datasets, but those that can bring together various expertise and make the wisest decisions. With their ability to evaluate value, ensure accountability, and provide an objective perspective, finance teams play an indispensable role in ensuring that AI investments yield real returns. By involving finance teams early and frequently, companies can transform AI from an exciting possibility into a reliable growth engine that can both drive revenue growth and achieve operational excellence.

Jason McDannold, Hoyoung Pak, Paula Walworth | By

Jason McDannold is a partner and managing director at AlixPartners and co - head of the Americas private equity and investor business. Hoyoung Pak is a partner and managing director at AlixPartners and global co - head of the firm's AI and data business. Paula Walworth is a partner at AlixPartners, working with companies to drive operational transformation and revenue growth based on technology investments.

This article is from the WeChat official account "Harvard Business Review" (ID: hbrchinese). Author: HBR - China, Translated and Proofread by Wu Qingya. Republished by 36Kr with permission.