Pick battles big enough to matter and small enough to win.
Many companies set their digital and AI transformations up to struggle from the start by getting the scope of the change wrong. Some companies start too small, believing an incremental approach will lower risk. This is a mistake. Successful transformations need to change something meaningful in the business, where there is a noticeable amount of value at stake and the impact can be measured. You won’t get far transforming a house by repainting the living room; you need to take on something much more substantial, like remodeling the kitchen.
Others, with the best of intentions, go too big too soon, and try to transform the whole company all at once. This is generally too disruptive, too expensive to do right, or too difficult to tackle as a first project, and usually fails. More commonly, companies spread bets and resources too thinly across an uncoordinated set of activities and initiatives. This results in a lot of activity but not much value.
The domain-based approach
The right approach is to identify a few important and self-contained domains in the business and rethink them completely. As many as 80 percent of successful interventions in a struggling digital transformation are based on re-anchoring the scope to drive a concerted effort against a well-defined domain.1 Taking this approach starts with identifying what the domains are. A domain is a subset of your enterprise that encapsulates a cohesive set of related activities. There are a few ways to define domains (Exhibit 1).
Companies can determine for themselves how best to draw a circle around a set of business activities to define a domain that makes the most sense. The key is to define a domain that’s large enough to be valuable and noticeable to the company, but small enough to be transformed without being unduly impacted by dependencies with other parts of the business. How many total domains are there for a company? The right number is about 10–15 for a monoline business. For a conglomerate, the right unit of analysis will be the strategic business unit, so domains are defined at that level. For the purposes of a digital transformation, however, select two to five to focus on transforming first. It is possible to go bigger and do more domains at the beginning. However, doing so requires a significant near-term investment, more coordination, and more talent. It may also carry more risk and will almost certainly require significant external resourcing and may deprive the organization of some early learnings. So be thoughtful about which and how many domains to tackle.
Creating value beyond the hype
Prioritizing domains
Prioritizing which domains to go after first requires an assessment along two broad dimensions: value potential and feasibility (Exhibit 2). Executives will recognize this simple way of prioritizing opportunities, but they should pay attention to the criteria that drive the assessment.
At this stage, a high-level estimate of the value potential based on a combination of outside-in analysis and discussions with senior leaders and industry experts will suffice. Most companies struggle with this kind of estimation because they lack experience understanding what might be possible with digital. To address this issue, consider using benchmarks from successful companies (even outside of your sector). Key value considerations include:
- Customer experience. Improving the customer experience should be considered the “first among equals” when it comes to considerations. Most successful digital and AI transformations are centered around the customer and meeting their needs. It helps to benchmark the current experience versus competitors, and project how much the experience would improve once the domain is transformed. This should translate into measures of specific customer satisfaction improvement as well as of customer growth and improvement in net value per customer.
- Financial benefits. At this stage you want to estimate financial benefits by focusing on operational key performance indicators (KPIs), such as new customer growth, reduction in churn, increased value per customer, improvement in process yield, or reduction in cost to serve. It can be difficult to precisely estimate how much improvement could be achieved at this stage, so an estimate informed by what others have achieved in analogous industries will suffice. Be careful not to downplay the potential at this stage. This is only for prioritization purposes and it is not yet the business case.
- Speed to value. A domain-based transformation should typically generate significant value within six to 36 months, depending on the domain. This is often an important consideration in terms of providing early benefits and helping fund the transformation. We have found that, in general, AI-intensive opportunities pay off more quickly.
- Synergy. If you are transforming more than one domain, synergy between them is a compelling point of leverage. Synergy can be evaluated in terms of three primary elements: data reuse across solutions, tech stack reuse across solutions, and shared change-management efforts. For example, if you develop a new mortgage sales platform and a new credit-card sales platform at the same time, you would only need to retrain your thousands of branch sales representatives once.
Assessing feasibility is a combination of understanding technical and data readiness, the change-management effort required, and the domain leadership’s capacity to commit. The most important considerations are:
- Strong executive sponsorship. You need to be clear whether the executives in the given domain are fully on board. A unit may be ripe for digital transformation, but if there are competing priorities, such as implementing a new IT system or completing a major compliance remediation effort, then it might not be the right time to transform it.
- Data and technology readiness. On the data side, the core questions to assess are the ease of moving the necessary data fields to the cloud and the quality of the underlying data. At this stage, a cursory analysis is needed, but more diligence will be required if this domain is selected. On the tech side, the core questions to assess are quality of the cloud architecture, the performance of the underlying core systems, and the ease of accessing data and applications with application programming interfaces (APIs). Your enterprise architects will be best positioned to make this assessment. Note that legacy technology—or the demands on large existing core systems, such as an ERP upgrade—is often used as an excuse for not making progress. These issues, of course, need to be understood, but are not a reason for inaction. Legacy mindsets are a bigger challenge than legacy technology.
- Ease of adoption. By understanding the scope, intensity, and risks involved in the change effort, companies can identify potential obstacles to the adoption of a digital solution. For example, carrying out change in a unionized environment may involve negotiations that can require time to carry out effectively.
- Ease of scaling. Assuming you successfully develop the digital solution, assess how easy will it be to scale it across the enterprise? How complex will the change-management challenge be? How many different data environments will the solution be operating in? These questions matter in getting the full value.
The output of this value-feasibility analysis should highlight two to five domains to prioritize. What’s important at this point is that you are not looking for precision—it does not matter if the measurements are exactly right at this point. View this more as a way to structure the conversation with the management team. Refining estimates will happen in the next step as part of reimagining the domain.
Rewired: The McKinsey Guide to Outcompeting in the Age of Digital and AI
In their words: Avoiding fragmentation and collaborating better
The enemy of any go-to-market digital transformation is fragmentation, especially in a company of Sanofi’s size. . . . You need to ruthlessly prioritize, and recognize that six months down the road, new shiny objects could creep in to dilute your initial objectives and slow down your ability to create big wins.
Today, we invest less in aggregate than we did three years ago, yet for the chosen priority projects, we devote more resources. Our iterative agile build cycles go faster and involve users in the development process, resulting in solutions that are much more relevant and impactful.
A second obstacle to success is us, the leadership, the management team. Everybody likes to have their sphere of business power, which in the past was linked to a specific, often siloed, P&L. The digital future is not like that. You need to be more open, able to empower, delegate, and collaborate. Nurturing digital capability is another factor. We had to train up a critical mass of people who could understand digital in order to attract and retain digital talent. Previously, we didn’t move fast enough and rejected innovative ways of working introduced by the new talent, which led to frustration and rapid churn.
Dr. Pius S. Hornstein, Global Head, Digital Global Business Units, Sanofi
In some cases, management will have a clear view of where the value is and decide to go after that domain immediately, skipping the prioritization step (see an example of this in the Freeport-McMoRan case study). This can be a good approach when management is clearly aligned and the value from the domain is meaningful. In practice, this can also be a useful way to build conviction in the organization by clearly demonstrating the value companies can capture or protect with digital and AI.
One large agricultural company decided to go this route by focusing initially on its commercial domain by supporting its agronomists in better serving its growers (customers) and in making it easier for these growers to do business with the company. The CEO and the top team were experiencing competitive pressure from new digital entrants and felt there were a number of customer pain points they could address quickly to improve cross-selling and retention.
While jumping quickly to address a trial domain can work well, leadership needs to guard against just launching another pilot that generates interest but doesn’t materially transform the business. That’s why it is so important to take the time to conduct a thorough domain reimagination effort.
Excerpted with permission from the publisher, Wiley, from Rewired: The McKinsey Guide to Outcompeting in Digital and AI by Eric Lamarre, Kate Smaje, Rodney Zemmel. Copyright © 2023 by McKinsey & Company. All rights reserved. This book is available wherever books and eBooks are sold.
