Idea in Brief
Established companies spend billions trying to turn themselves into digitized orchestrators of some new ecosystem, only to fall flat on their faces.
Why It Happens
The CEOs believe that the existential threat posed by digital disrupters requires a gigantic, model-busting response.
Adopt an incremental experimental approach: discovery-driven digital transformation. Look for problems to fix with digital technology, but exploit your rich knowledge of customers, broad operational scope, and deep talent pools while learning your way to a new business model.
What’s your digital strategy? That simple question often throws the CEOs of traditional companies into a panic. They believe that digital technologies and business models pose an existential threat to their way of doing business—and of course they’re right. But the pressure they feel often leads them to make big bet-the-farm moves—and that’s usually wrong.
Veon, a large multinational provider of telecommunications services, is a case in point. Its new digital platform, introduced in 2017, was a huge project, involving 100 staff members in Amsterdam and another hundred or so in its London office. The idea was to create a mobile app that would offer users rich localized experiences and serve as a sales channel for Veon’s commercial partners (such as Mastercard). Management considered the project its top priority. But after being launched with much fanfare, the app got a lukewarm response from customers, and the effort to build a new ecosystem around it was scrapped. The failure led to a management exodus, layoffs, and a back-to-basics strategy with digital efforts sidelined to pilot-project stage.
Veon still needs a new business model, though, and clearly can’t afford to make many more large investments in searching for one.
It doesn’t have to. Just because a threat is huge doesn’t mean that a response has to be. To the contrary, companies like Veon would actually be much better off taking a more incremental approach to transformation over time. While they should always have a vision of where they want to go, they should work their way toward it by continually finding opportunities to digitize problematic processes in their core operations. When they tackle those projects, they’ll learn what metrics to use, which assumptions to revise, where they can introduce new business models, and who their new competitors might be. And as they absorb those lessons, their understanding of their competitive landscape—and the long-term goals they set for themselves—will inevitably change.
There’s already a process for this kind of ongoing learning approach to strategy: discovery-driven planning (DDP). One of us, Rita, and Ian MacMillan developed it in the 1990s as a product innovation methodology, and it was later incorporated into the popular “lean start-up” tool kit for launching businesses in an environment of high uncertainty. At its center is a low-cost process for quickly testing assumptions about what works, obtaining new information, and minimizing risks.
In the following pages we’ll describe how an adapted form of DDP can help incumbent firms confront digital challenges and learn their way toward a new business model. Let’s begin by looking in more detail at why a step-by-step transformation works better for traditional firms than the all-or-nothing approach that characterizes a start-up’s pivot.
The Incremental Advantage of Incumbent Firms
Economists have long puzzled over why firms exist at all and, at a more granular level, which tasks belong within the boundaries of a given firm. One line of thought, begun by Ronald Coase in the 1930s, suggests that under certain conditions, market transactions often are not satisfactory for individuals: when it is difficult or expensive to get information about what you want to buy, when bargains are hard to strike because information is asymmetrical, and when it’s costly or challenging to enforce agreements. If any of those conditions apply, it makes sense to keep the activities involved within a firm.
Until fairly recently, the boundaries between firms and markets were well understood and relatively fixed. But digital technologies have changed all that by making it possible to use markets for a lot of work that once was done more efficiently within firms. Platforms such as Alibaba and Amazon have made it easy to outsource functions like selecting suppliers, negotiating prices, enforcing contracts, managing payments, and more.
As a result, executives in companies that were born digital have assumptions about how transactions should be structured that are completely different from those of executives in legacy companies. What’s more, because digital firms’ structures are evolving all the time, their managers revisit those assumptions frequently. Direct-to-consumer businesses (think Casper in mattresses, Harry’s in shaving, and Warby Parker in eyeglasses) are constantly experimenting with and adjusting features like free shipping, product bundles, bonuses for adding items, and so on. Those tactics simply aren’t available to an incumbent selling through distributors. And because the digital businesses cut out intermediaries, they can be profitable at a much lower scale.
A key consequence of all this is that digital start-ups can change direction, or pivot, without destroying much value. They usually aren’t that capital-intensive and don’t have big payrolls. The founders of Rooted, for instance, initially sold plants out of their apartment directly to consumers, only later moving to a separate space and hiring employees. For such companies, failure is relatively cheap—unless it happens late in the day (or investors succumb to the growth-at-all-costs mantra that is unraveling the fortunes of many so-called unicorns).
The employees, managers, and shareholders of traditional companies, however, cannot pivot without destroying value. If their digital gambles fail, workers lose their jobs, and physical assets have to be unloaded at fire-sale prices. And unlike the venture capitalists who back start-ups, the investors in what was once a safe company may not have the buffer of high-return investments to offset their losses.
But although incumbent firms can’t pivot easily, the good news is that they don’t need to. Think about what big companies can do that start-ups can’t. Entrepreneurial ventures nearly always exploit a single idea. They usually can’t try out multiple versions of the same idea at the same time, let alone multiple ideas. A big firm, in contrast, has the resources to explore a variety of ideas and can more easily experiment with different processes and operations, which makes it more likely to discover a dominant model than a start-up is. This also gives a large firm a better chance of responding effectively to a digital challenge.
Goals should frame the technology as an opportunity for the business.
Take the case of the German metals distributor Klöckner. Its CEO, Gisbert Rühl, wanted to build a digital platform for the entire industry—but he didn’t sponsor a big-bang effort to create one. Instead, his goal was to build digital competencies gradually, while benefiting from the knowledge and insight of people working in the firm’s core steel business. For the first two years Rühl focused on digitizing inefficient manual processes; the firm created an online shop, a contract portal, order transparency tools, and a parts-manager app. Through these efforts it learned enough to create a platform on which the company and customers could seamlessly interact.
Klöckner’s story reveals another advantage that incumbent firms have, at least in the early stages of an industry’s adoption of digital models. They’re led by people who already know their customers and can mine rich databases of prior transactions for insights. Start-ups are often led by technical experts and tend to be driven by new technical functionality rather than by the full portfolio of what customers are looking for. If you put a team of people who know the customers on the job, you’ll stand a better chance of making your digital investment pay off. That’s why Klöckner insisted that every project focus on how to help customers communicate more easily and efficiently with the company. That isn’t the only goal to set, of course. Another company might start with a priority on shortening the time it takes to respond to a customer request. But whatever the goal is, it should frame the technology as an opportunity for the business rather than frame the business as an opportunity for the technology.
Once you accept the idea that firms should aim to disrupt in a nondisruptive manner, the challenge is subtly transformed from “What new business model should we back?” to the more nuanced question, “How can we learn our way toward a model that’s right for our business?” That is where DDP comes in.
The Digital Context
DDP is somewhat like reverse engineering. When you use it in product development, you begin by imagining the offering you want to create and then figure out what you would need to change in order to get there. When you apply it to digital transformations, however, the focus is on reinventing the way you sell and deliver the products you already make as well as on identifying how to create and deliver new value through new digital capabilities.
Take power generation. Digital technologies are disrupting this once-stable industry, just as they are many other industries. Traditionally, power was generated from a central source and sent to its destination over a centrally managed grid. But new advances have made it possible to dynamically distribute power generated from dispersed small-scale producers tapping multiple energy sources. People with solar panels on their roofs or windmills in their gardens can sell surplus energy back to the grid, making households’ cost of investing in power generation hardware more affordable and reducing the public’s reliance on huge fossil-fuel power plants. If incumbents assume that the old business model will predict future success, they’re likely to make big mistakes. General Electric’s failed bet on the continued dominance of fossil-fuel-based electric plants provides a dramatic example.
Let’s explore what’s involved in applying a DDP approach to digital transformations. There are five key steps:
1. Define the Operating Experience: It’s Not Just About Digital
Before investing in a line of code, look for what isn’t quite working in your operation. Where do you regularly need workarounds or have to stop a process unexpectedly to fetch more information or involve another person? These are likely to be areas that digitization can improve. Then think about how to redesign your operations there so that technology adds value, by making offerings and processes better, faster, cheaper, or more convenient.
The retailer Best Buy is one incumbent that was able to reconfigure its business operation in a way that created competitive advantages the digital-only players couldn’t replicate. Back in 2010, Amazon released its price-comparison app, one of many tools that allowed shoppers to check out products in a physical store but order the same items at a discount online. Called “showrooming,” the practice threatened to squeeze the lifeblood out of retail chains, which struggled to offer competitive prices while paying for real estate, staff, and inventory. It was one of the reasons Best Buy lost $1.7 billion in a single quarter in 2012.
Hubert Joly, the CEO hired to turn the company around, centered his strategy (and his business model) on solving two problems: negative comparable sales and declining operating margins. To do this, he envisioned a company that blended the human, the physical, and the digital in ways that an online-only player would find hard to match. He began by imagining what kind of customer experience Best Buy could deliver and, more important, identifying where it hadn’t leveraged digital technologies to create that experience.
From this was born Best Buy’s Renew Blue project, which had five components: a reinvigorated customer experience; a change in vendor partnerships; investments in ecological and social initiatives; the employee experience; and a return for investors. Financial targets and experiments were set up for each component.
To improve the employee experience, Best Buy launched initiatives focused on workforce morale, such as bringing back a popular employee discount that had been discontinued and investing in more-intensive training. To appeal to customers, the company began to match the prices of Amazon and other e-commerce players, which required a massive effort to overhaul Best Buy’s warehousing, software, and supply chain activities. But because customers could walk out of the stores with the products, they could avoid the wait and the hassles (such as porch piracy) of having expensive products delivered, and that gave Best Buy an important edge. The company also created a system through which customers could order goods online for delivery or for pickup at the store. With 70% of Americans living within 15 miles of a Best Buy outlet, that approach proved to be extremely cost-effective.
Best Buy’s new model turned the disadvantage of costly real estate into an advantage. At its more than 1,000 big-box locations, brands such as Apple, Samsung, and Microsoft created stores-within-stores, essentially paying rent to feature their offerings where real live shoppers could discover them in person. Best Buy is a neutral party to warring tech giants; archrivals Amazon and Google both sell their goods there. Finally, Best Buy invested in an in-home adviser capability, in which salaried, highly trained consultants go to customers’ homes and provide tech help without selling anything. The goal is just to build stronger relationships with consumers. Throughout it all, Best Buy steadily transformed its digital footprint to support the strategy.
The Best Buy story illustrates the importance of being willing to rethink assumptions about how to use assets and engage with partners. Previous leaders in the firm had failed to see any way that it could price-match online retailers. But because Joly challenged traditional thinking, he spurred the company to reimagine relationships with vendors (which now pay to be in Best Buy stores) and redesign its supply chain so that the company’s physical assets could support a new business model for competing with e-commerce giants.
2. Focus on Specific Problems: Identify Outcomes and Progress Metrics
The key question in any digital-transformation strategy is, How can we use data and digital capabilities to create new value for our customers? The DDP process translates that challenge into clear project goals.
A traditional success metric for new projects, even today, is return on investment. But ROI doesn’t help you understand what value a project adds for customers, at least not directly. Further, to calculate it you need to estimate both investments and returns, which is precisely what you haven’t figured out yet. What you need to do instead is identify metrics that are more closely linked to the specific improvements you hope digital initiatives will bring about.
We typically collect all this information in a “from-to” table, which identifies a problem, describes what a solution would achieve, and proposes a way to measure progress on that solution.
As you work through solving these problems, you’ll test and refine your assumptions—a key DDP discipline. You can also capture what you’ve spent to gain new insights and what they’ve saved you. Eventually, you can back into something similar to an ROI calculation.
A key part of discovery-driven transformation is identifying organizational problems that can be addressed with digital technology, the desired improvement for each, and a metric for assessing progress toward it. All this information is captured in a tool called the from-to table. Below is how one financial services organization we’ve worked with filled its table out.
|No consistent information about investments in a portfolio of projects; manual process||Clear and easily obtained information about investment flows; automated process||Reduction in time needed to update portfolio review information from 10 days to seconds|
|Significant effort needed to onboard new team members and bring them up to speed||Automated onboarding assistance that helps new team members learn the background of a project||Reduction in time it takes new team members to reach productivity from 30 days to five; high engagement scores among 85% of team members and top quadrant scores for psychological safety|
|No capture of learning created in one part of the organization for reuse elsewhere||Routine recording of project insights in a database that’s searchable by keywords, geographies, and contexts||Information is shared by an average of 10 units in the organization; 50% decrease in duplicative experiments|
|Slow and inflexible financial and talent resource allocation across new opportunities||Dynamic prioritization and resource allocation driven by real-time data and discovery||Resource reallocation cycles go from quarterly or annual to weekly. Annual 50% to 100% increase in number of experiments with strategic options|
At Klöckner, the ultimate goal was to change the business model in steel from marking up inventory to a services revenue model. At first, the digital initiatives were simple and were focused on improving the order process—by, for instance, replacing the faxing of orders with an online portal for ordering. With each one, performance on metrics such as turnaround time and the number of steps required to complete an order improved. As the company gained more knowledge and capabilities, its projects became more ambitious.
Of course, you still need a way to measure progress on digital transformation overall, and to do that we suggest a metric we call return on time invested (ROTI). To calculate it, you simply divide your total revenue by the number of employees. The idea is that successful technology investments should let you accomplish more with fewer people. For example, we used annual report data from 2018 to compare Amazon (a digital-first company) with Walmart (a more-traditional legacy business). We found that Amazon had $232.9 billion in net sales and 647,500 full- and part-time workers. Its sales per employee were $359,671. In contrast, Walmart had $495.8 billion in net sales and 2.3 million associates. Its sales per employee were $215,548. Amazon enjoyed 67% higher performance per employee.
3. Identify Your Competition: Cast a Wide Net
Industry boundaries have blurred so much that standard industrial classification (SIC) codes are more or less useless. This by itself is one reason why conventional strategy-making approaches predicated on boundary assumptions are failing incumbents.
We suggest that leaders instead think about the field of competition not as a marketplace where similar players offer rival products and services but as what strategists call an arena. An arena is defined by a customer need—what Clay Christensen dubbed the “job to be done.” It’s a notion that goes back to Ted Levitt, who recommended that railway companies see themselves as competing in the transportation business against airlines, buses, trucking, and even cars. If railway passengers are a market, transportation users constitute an arena.
Smart born-digital firms already think this way. For example, Netflix has been very clear that it doesn’t intend to compete just against television or the movies for viewers’ time. It intends to compete against every possible leisure activity that a person might do instead of watching streaming content. The company sees traditional media companies as its rivals, of course, but its leadership looks at magazines, books, podcasts, and sporting events as competition as well.
At this point in the process, you should go back and determine whether the outcomes and metrics of success you spelled out in steps one and two are reasonable, given the arena you’re competing in. Is your category losing share of wallet to others in the arena, for instance, or holding its own? Netflix has plenty of room to meet its growth goals, because total hours of video viewing are increasing and a lot of that growth is from streaming video.
4. Look for Platforms: Don’t Forget the Ecosystem Implications
In the digital economy, striving to become an intermediary through which others buy and sell goods is an extremely popular strategy. It’s a tempting business model, because once the two sides of a market have joined a platform they have little incentive to jump to another. This is partly due to network effects, whereby a platform’s value to any user increases as the number of other users on that platform rises. Airbnb, for instance, benefits when more hosts and more guests use it and has historically gone to great lengths to ensure the loyalty of both.
ROI doesn’t help you understand what value a project adds for customers.
A platform is also attractive because it needs less capital. To run a conventional hotel, you have to have real estate, rooms that need to be looked after, reservation systems, staff, and so on. Airbnb, in contrast, taps an ecosystem of hosts to provide all those things, and its directly controlled activities are simply to match hosts and guests and guarantee transactions, both of which occur entirely in the cloud and thus are infinitely scalable.
To understand whether a platform opportunity exists, we use a tool we call a customer consumption chain (introduced in HBR in 1997). The idea is simple: that as customers try to get jobs done in their lives, they go through a series of experiences, beginning with awareness of a need, then working through how to get that need met, and going all the way to the conclusion of a service or the end of a product’s life. Digital technologies make open-market transactions for many links in that chain possible, allowing firms to build platforms.
That sounds like bad news for established organizations. But they have an ace in the hole: They employ many people who have deep technical expertise or understand customer problems. Those capabilities can give them an edge in identifying platformlike opportunities and building ecosystems. At Klöckner, Rühl realized that once there was price transparency—and far less friction—in the trading of basic metal commodities, competitive advantage would shift to suppliers that could offer superior solutions and service. The company blended the new ways of operating on platforms (co-creating designs with customers, for instance) of its digital arm with its workforce’s deep expertise (in, say, manufacturing with 3-D lasers) to develop customized, higher-value offerings.
Becoming a popular platform isn’t easy for corporations. The business landscape is littered with would-be platforms that failed even though they seemed to have all the right components. General Electric’s Predix initiative, which was intended to be the platform for the industrial internet of things, is an example. Rather than driving the digitization of services that customers would value, Predix was sucked into serving primarily internal GE units—and a lot of them. Further, as part of GE Digital, the initiative was given P&L responsibility, which oriented it toward short-term contracts with customers that could pay some bills in the interim. It also took on way too much too soon, rather than proceeding by finding a good fit for its capabilities and building from there.
5. Test Your Assumptions: Failures Are Lessons Too
One of the more popular tools to come out of DDP is the assumption checkpoint table. To create one, just write down the next few milestones that your digital project will go through, which assumptions need to be tested at each, and if possible, how much that test will cost. The beauty of this approach is that it moves the conversation from “Oh, you were wrong, that was a failure” to “Was it worth that price to learn what we needed to learn?”
Consider how Buffer, a service that allows people to space out social media promotions without having to predetermine the timing, tested assumptions in its launch phase. Joel Gascoigne, Buffer’s cofounder, got the idea for the business from his own frustration with how clunky it was to try to tweet more consistently.
The first assumption he wanted to test was whether anybody else perceived this to be a problem. So he built a very simple two-page website. The first page’s pitch was “Tweet More Consistently with Buffer.” If users clicked on it, they were taken to a second page, with the heading “Hello, You Caught Us Before We’re Ready,” which had a place for people to enter their email addresses if they were interested in Buffer’s solution. Most people weren’t, but some were. So Gascoigne added a third page between the other two to test pricing hypotheses. And again, most people weren’t interested in paying, but enough were to persuade Gascoigne to build the product.
Next he had to decide how complex to make it and how many social platforms to apply it to. He ended up keeping it very simple and supporting only Twitter at first. As of 2018, Buffer had more than 1.4 million social accounts connected to its apps.
Many large corporations have adopted a similar test-and-learn mindset. Several new services make experimentation easier—for example, Alpha, whose subscribers use it to obtain fast feedback about products from potential customers before making expensive or irreversible decisions. At WellMatch, an Aetna business unit, experimentation helped resolve disagreements about design decisions. According to Etugo Nwokah, the former chief product officer, one area of disagreement involved its website: Every group in the unit wanted to have its content appear on the landing page. The trial entry page ended up being so busy that it confused consumers. The company had to go back to the drawing board and do a redesign—but was able to do so at a much lower cost and risk than if the webpage had been launched for real.
Digital transformation is complex and requires new ways of approaching strategy. Starting big, spending a lot, and assuming you have all the information is likely to produce a full-on attack from corporate antibodies—everything from risk aversion and resentment of your project to simple resistance to change.
A discovery-driven approach gets leaders past the common barriers to digital transformation. By starting small, spending a little on an ongoing portfolio of experiments, and learning a lot, you can win early supporters and early adopters. By then moving quickly and demonstrating clear impact on financial performance indicators, you can build a case for and learn your way into a digital strategy. You can also use your digitization projects to begin an organizational transformation. As people become more comfortable with the horizontal communications and activities that digital technologies enable, they will also embrace new ways of working.
Incumbent companies have some great advantages over new competitors: paying customers, financial resources, customer and market data, and larger talent pools. But CEOs will have to integrate agility and innovation into their broader organizations and communicate the new ways of digital thinking while minimizing disruption to their existing businesses. A discovery-driven approach provides a way to address those challenges.