How Should You Value AI and Digital Initiatives?

Like other initiatives to improve performance, AI and digital projects fall into one of four categories based on their intended benefits: cost reduction, improved customer experience, new revenue sources, and better decision making. Here’s how to think about the value of investing in those initiatives...

In recent research by McKinsey & Co., 42 percent of chief strategy officers said they were struggling to drive impact. And AI is making that pressure worse: fellow C-suite leaders and board members want to know what all the new investment is producing.

In the new book Strategy and Value: Mastering the Decisions that Matter, co-authors Tim Koller, Andy West, and Whitney Zimmerman argue that the answer isn’t a special playbook for AI; it’s holding AI to the same principles that should govern any big bet.

Strategy and Value.jpg
Excerpted from the new book Strategy and Value: Foundations That Drive Long-Term Performance

The book, written by two McKinsey partners and one of the firm’s strategy experts, is built on four foundational ideas. Two come from finance: a company’s value is driven by its return on invested capital (ROIC) and revenue growth, not by earnings growth alone; and managing for long-term value beats chasing short-term earnings (routinely cutting costs at the expense of growth, for instance, tends to destroy value).

The other two come from strategy: companies need a shared understanding of why their markets are becoming more or less attractive, so they can put resources behind their best opportunities; and they need competitive advantages that are unique and durable, not capabilities that rivals already have or can easily duplicate.

In the excerpt below, the authors apply that thinking to AI and digital initiatives.

•••

Applying Valuation Principles to AI and Digital Initiatives

The transformative effects of technological developments such as artificial intelligence (AI) and digitization are top of mind for executives and boards. For most companies, AI and digital are technologies to help improve operational and commercial performance. The technology does not change what these companies sell, so much as how cheaply and well they can sell it. In some cases, new companies will use AI and digital to disrupt a market entirely.

Sometimes executives argue that AI and digital initiatives are “strategic” and that their benefits can’t be measured. In our experience, this is rarely the case. The error in this line of thinking originates in an improper definition of the base case. Ideally, managers should analyze all investment decisions compared with an alternative course of action. For AI and digital projects, the alternative may be to do nothing. But the do-­nothing case doesn’t mean a company’s cash flows necessarily remain the same.

Banks have faced this challenge several times. In the 1970s and 1980s, banks introduced automated teller machines. In the 2000s, banks set up online banking. In the 2010s, banks developed mobile banking apps. In hindsight, it seems obvious that banks needed to introduce these innovations. But because every bank adopted them, it was difficult for any single bank to gain a lasting revenue advantage.

Thus, although a mobile banking app makes strategic sense, it appears to create a negative present value due to its additional costs and the lack of added revenues. However—­ and this is the importance of the do-nothing case—­ if the bank doesn’t build a mobile app, it will likely lose market share and revenues. In this case, the cash inflows are the avoidance of lost revenues, potentially a substantial amount. Then this “strategic” project likely does have a positive present value after all.

Like other initiatives to improve performance, AI and digital projects fall into one of four categories based on their intended benefits: cost reduction, improved customer experience, new revenue sources, and better decision making. 

Like other initiatives to improve performance, AI and digital projects fall into one of four categories based on their intended benefits: cost reduction, improved customer experience, new revenue sources, and better decision making. In some cases, the impact could have multiple benefits. For example, in pharmaceuticals, AI could help companies develop novel drugs and speed up overall drug development while lowering its cost. In transportation and logistics, automated vehicles, along with sophisticated scheduling and better understanding of market conditions, could improve pricing, lower costs, and enhance on-­ time delivery.

Cost Reduction

Many AI and digital performance initiatives can help companies reduce operating costs. For example, predictive maintenance on factory equipment can cut both maintenance costs and lost production from downtime. Another example is robotic process automation. This term doesn’t refer to physical robots but rather to software that automates processes such as accounts payable. As this software, including AI tools, becomes more sophisticated, it can take on even more difficult tasks, such as handling billing exceptions.

Using process automation, one mining company saved more than $360 million per year. It gave managers more insight into what was happening in the field, enabling them to make production adjustments and anticipate maintenance and procurement needs.

In another example, fossil fuel power generators have improved how efficiently a plant burns fuel by up to 3 percent by using sensors and actuators for remote monitoring and automated operations as well as smart valves that self-­ report and repair leakages. These power companies have also used automated generation of work orders, remote expert support employing virtual reality devices, and automated warehouses to reduce operating costs by 5 to 20 percent.

If competitors are pursuing AI and digital initiatives to reduce costs and you are not, eventually you’ll still face pressure to reduce prices in line with competitors. 

The do-­nothing case becomes more important to consider in situations where savings from cost reductions are likely to be passed on to the customer as price cuts. If competitors are pursuing AI and digital initiatives to reduce costs and you are not, eventually you’ll still face pressure to reduce prices in line with competitors. The alternative to pursuing the digital initiative would be a decline in cash flows due to lower prices without reduced costs. The present value of the initiative may turn positive when you compare your initiative with the relevant do-­nothing case.

Improved Customer Experience

Consumers are already benefiting from companies’ AI and digital initiatives. Many retailers have become omnichannel, giving consumers shopping flexibility. Consumers can purchase clothing in a store or online and have it shipped to their homes or to a local store. A customer who decides to return an item can do so at any store or mail it back, regardless of how it was purchased.

Using AI and digitization to improve customer experience can add value to a business in various ways. One leading manufacturer of agricultural products was struggling with low customer satisfaction scores and an erosion of its customer base. Employing digital solutions, the

company created a seamless online process for ordering, tracking, and query management. This increased the company’s customer satisfaction score by 24 percentage points and improved throughput by 20 percent.

In some cases, improved customer service also reduces costs. An electricity distribution company redesigned its customer interfaces in a digital-­first way that prioritized the customer’s online interaction. Customer satisfaction rose 25 percentage points, employee satisfaction increased by 10 percentage points, and customer service costs fell 40 percent.

As with applying AI and digital solutions to reduce costs, it’s critical to think through the competitive effects of investing in digital to provide superior customer experience. Recall our earlier example of the mobile banking app. The value proposition boils down to cash flow, but special considerations emerge. Does improved customer service lead to higher market share because your customer service is better than that of your competitors? Or does it maintain your market share or avoid losing market share because competitors are doing the same thing? In many situations, a company that adopts an AI or digital initiative of this kind will see its competitors embrace similar approaches, leading customers to expect an improved experience without paying extra for it.

New Revenue Sources

Some companies have generated new revenue sources through AI and digital initiatives. Consider Itaú Unibanco, the largest bank in Brazil. Itaú has launched innovative financial products that expand services for existing clients while attracting new ones. These offerings include an AI-­powered WhatsApp interface for the Pix payments system and peer-­to-­peer transfers, an AI-­enabled wealth advisor, and an AI-­driven financial management solution for small- and medium-­size businesses.

In a similar way, Ford offers software services to its commercial vehicle owners to monitor and analyze in real time their fleet locations, efficiency, and maintenance needs. Another transportation company’s digital solution helped its customers improve fleet maintenance, generating more than $10 million in revenues.

Better Decision Making

Some executives are using advanced analytics and AI techniques to enable better decision making. For example, a maker of high-­tech hardware improved pricing for thousands of product configurations through automation. Key features of the approach included price benchmarking, analysis of price trends, and automated pricing recommendations with weekly updates of up to 200,000 price points for as many as 20,000 products. 

In another case, a consumer products company used advanced analytics to improve the design of its planograms. A planogram is a model of how a consumer packaged goods company allocates its limited space on retail shelves. It describes which products will be included and how to display them. Analytics showed managers at the company how they could improve sales. At the same time, the company reduced the number of people needed to design planograms from 10 to two.

Advanced analytics to improve decision making can generate additional revenues, reduce costs, or both. In the planogram example, the improvement can increase total customer spending by getting customers to upgrade to more profitable products. Because the change involves only choices within the company’s product mix, the improvement can create value without necessarily inviting a competitive response. In other cases, the benefits may be diluted because competitors take similar actions, but the investment in analytics still may create value by maintaining competitive parity.

...New companies and business units of established companies are also being created primarily to provide AI and digital products and services.

As we have seen, companies are deploying AI and digital solutions to improve the performance of existing businesses. But new companies and business units of established companies are also being created primarily to provide AI and digital products and services.

These include software companies such as OpenAI and Anthropic or new data center companies such as Prometheus Hyperscale. Meta has formed several businesses, including one for AI-­enabled wearables. AI and digital businesses typically focus on providing services such as AI and other software, electronic payments, content streaming, e-­commerce, social media, marketplaces, and mobility and delivery platforms. For example, Zoom Video Communications has largely replaced traditional telephone conferencing services, and the company’s revenues have increased from $60 million in 2017 to more than $4.5 billion in 2024.

Digital businesses can also operate in multiple categories, although the evolution is not easy to carry out. Amazon shows how it might be done. It’s widely known that Amazon started as an online retailer.

Over time, it expanded into a marketplace that allows third-­party sellers to list products on Amazon’s website. Amazon earns fees from these sellers, including commissions on sales and optional fees for services such as warehousing, fulfillment, and delivery. Amazon then used its expertise in running its own data centers to become a leader in cloud-­based computing (Amazon Web Services).

Although still speculative at the time of writing, AI could disrupt entire industries such as enterprise software (if customers can develop in-­house tools for some tasks), legal and accounting firms (where AI is already reviewing contracts and analyzing accounts), media businesses (where AI might develop some content), and radiology and pathology in healthcare (where AI can already read various scans more consistently than humans). The extent of disruption likely will depend on how close the effect of AI is to the core business activities.

For established companies, we usually begin the valuation process with a company’s current performance and project into the future. For new AI and digital companies, we find it’s better to jump to the future and work back to today.

Valuing these new businesses is difficult because they are at such an early stage in their life cycles that historical performance is irrelevant. For established companies, we usually begin the valuation process with a company’s current performance and project into the future. For new AI and digital companies, we find it’s better to jump to the future and work back to today.

These new businesses have the same value drivers—­ return on invested capital (ROIC) and revenue growth—­ as established companies, so they can be valued using a discounted cash flow (DCF) model. However, there’s great uncertainty about how big these newer companies can get and what kind of margins and ROIC they can earn. So, the valuation process requires adaptation of the standard DCF model.

To that end, we develop scenarios about what the revenues, margins, and ROIC of these companies might be in seven to 10 years. We do this because the economics of the companies may be clearer then.

 Also, their future economic performance is much more important for the purposes of valuation than the performance path they take between now and the future.

This approach requires you to visualize what a business might look like in the future. What product or service will it be selling? How big will the market be? How competitive will the market be? How will that affect margins and returns on capital? Keep in mind, when developing scenarios, that you should consider both failure and success modes. History shows that many of these kinds of companies will fail. When estimating the size and ROIC of an early-­stage AI and digital business, carefully examine the underlying economics to determine whether the business might be a capital-­light or a winner-­takes-all business. Some new digital business models are indeed capital light, involving only a modest amount of property, plant, and equipment compared with the revenues generated. But do not automatically assume that all digital businesses are capital light. Many leading digital businesses have spent heavily on data centers or chip factories.


Excerpted with permission from the publisher, Wiley, from Strategy and Value: The Four Foundations That Drive Long-Term Performance by Tim Koller, Andy West, and Whitney Zimmerman. Copyright © 2027 by McKinsey & Company. All rights reserved. This book is available wherever books and eBooks are sold.


More like this
COMMENTS

Thanks to our strategic partners

AIPath logoAucctus logoDisruptive Edge logoEzassi logoHYPE logoITONICS logoPeer Insight logoQmarkets logoWellspring logo