Managing Director & Senior Partner
Related Expertise: Innovation Strategy and Delivery, Data and Analytics, Digital HR
This article is a chapter from the BCG report, The Most Innovative Companies 2018: Innovators Go All In On Digital.
In the innovation survey for this report, 30% to 40% of respondents said that they expect big data analytics and digital design to have a significant impact in their industries over the next three to five years. About a third said their companies are targeting data analytics in their innovation and product development efforts, and about a quarter are targeting digital design.
These figures are significant increases over those in the last survey, but given the impact of digital disruption across multiple industries, one might ask why the numbers are not even higher. Our experience suggests that a big part of the answer lies in the scale of the undertaking. Digitizing processes such as innovation programs requires the ability to access large quantities of data from multiple sources, the technology and skills to analyze the data and extract insights, and the process acumen to work in nontraditional ways, such as agile—all of which changes how companies go about innovation. It’s a daunting challenge, and most companies do not yet have the requisite capabilities.
The potential impact of data analytics and digital design on innovation strategy is a big deal. (See the companion article “How Digital Transforms Innovation Strategy.”) But the potential impact on operational processes, including R&D and new-product development, is just as significant.
Digital’s impact on operations generally takes three forms. The first is streamlining and speeding up processes that traditionally have been handled manually or are still paper-based. In the pharmaceuticals industry, for example, think about the ways in which data, mobile technology, and blockchain can revamp how companies identify participants for, and conduct, clinical trials.
The second, more far-reaching, impact is digital’s transformation of the process of innovation—in other words, R&D itself. Data analytics and other digital capabilities can handle tasks that humans cannot, such as processing massive amounts of data from disparate sources to find patterns that are otherwise hard to discern.
For example, medical researchers used data analytics to uncover the genetic patterns that underlie certain diseases. That information was then used to predict outcomes for drugs targeting the proteins associated with the relevant genes. This data-backed insight led to the discovery and development of PCSK9 inhibitors, a class of drugs that lower cholesterol. Or consider a financial services company that wants to shift from a products-and-process business model to one built on customer journeys. The company will need to adjust its products to create new digital offerings, such as online auto insurance, and digitize its systems and processes for effective product and service delivery.
These examples are just the tip of the iceberg. As capabilities improve, companies could end up reworking their entire R&D or product development value chains to take advantage of new ways of generating and evaluating insights—in many cases short-circuiting protracted, risky, and expensive steps in their current ways of doing business.
Digital natives often have the advantage of designing their digital innovation processes from scratch. This has led to disruptions in industries as varied as agriculture, consumer goods, manufacturing, and financial services, and it is requiring all companies to rethink their operational processes for innovation. For most established companies, the digitization of R&D and product development is a substantial task that needs to be approached with a transformational mindset.
One global automaker, which is at the forefront of using digital to reshape its R&D process, has established a digital center with the following mandate:
The third impact of digital on operations involves the tools that companies use to manage the portfolio of innovation opportunities. In our experience, more and more companies are adding automation to their portfolio management approach and digitizing their pipelines by using data and analytics to help prioritize ideas for development. It’s not unreasonable to expect that more advanced innovators will soon employ predictive algorithms that will tell them which ideas have a higher likelihood of success.
Regardless of industry, digitizing a large company’s product development processes takes time and effort. Companies need to start by harnessing data and adopting digital ways of working.
Data. Data is the fuel for the digital innovation engine. It can come from customers, processes, machine operations, testing, production plants, storage facilities, and delivery logistics systems, among myriad other sources. Companies need the ability to both access and process large amounts of disparate data—including data from third parties—on a continuous, reliable, and repeatable basis. But harnessing all that data is no simple task.
First, too much company data today is siloed. It belongs to the marketing, finance, or sales department, and that department is the only one that has access to it. Companies need to adopt an open-source approach so that the entire organization, including R&D and product development teams, can access data wherever it resides.
Second, old-style data warehouses limit the kinds of data that companies can collect and what they can do with it. Many companies are restructuring their data collection, storage, and usage approaches into data lakes—large repositories of data in a “natural,” unprocessed state. Because of their flexibility and size, data lakes allow for substantially easier storage of raw data streams, which today include a multitude of data types. Data can be collected and then sampled for ideas, tapped for analytics and feedback loops, and even potentially treated for analysis in traditional structured systems. While data warehouses typically provide backward-looking views, product development organizations need data to tell them not just what happened in the past but also what is likely to happen in the future. They want predictive and actionable insights to inform their R&D.
Third, companies have more opportunities to interact with customers and suppliers than ever before, opening up different ways to experiment with new products and services, learn what buyers want, and adapt accordingly. But most companies have so far not capitalized on this opportunity because they lack the capabilities to follow the customer’s digital trail, and they have not established the kind of customer feedback loops that allow for experimentation and a test-and-learn approach. Moreover, customer data often remains locked up in the customer insight function, never making its way into business decisions or product development programs.
General Electric, a constant presence on our list of the 50 most innovative companies and number 18 this year, is an exception to this pattern. Its FastWorks program is modeled on many of the practices used by startups to move new products quickly to market, including building customer feedback into the R&D process. FastWorks involves customers early on in the process and uses frequent testing to confirm or disprove assumptions and to guide adjustments throughout the development process.
Ways of Working. Digital innovation processes are cross-functional and increasingly agile. Digital skills will neither thrive nor be particularly useful when used in solitude. Companies need to find ways to encourage, or even compel, collaboration among people with digital and traditional skills and expertise. But this means that, just as digital experts need a working knowledge of the business, business people need to understand the basics of digital. Some equipment manufacturers have developed new revenue streams from service businesses that use digital technologies to maintain and support capital equipment. Siemens’s train engines are one example. The company’s development of predictive maintenance capabilities required more than digital knowledge of sensors, data, and the IoT; engineers also needed to know about the mechanics of train engines, how customers use them, and the economics of maintenance for complex machines.
Digital innovation processes are increasingly agile because agile ways of working are more collaborative and faster than traditional methods. Cross-disciplinary, collocated teams collaborate in innovative ways on the basis of insights gained from data and customer feedback. By working iteratively and incorporating feedback to improve continually, agile teams can transform innovation from the inside out. Because many companies are still organized around highly specialized functions, however, the shift toward agile often requires process redesign and organizational change—from large functional structures to small teams of cross-trained individuals. (See the companion article “Organizing for Digital Innovation.”)
Agile ways of working are particularly conducive to a test-and-learn approach, which is the hallmark of innovation for many digital natives. Rather than spend months or years designing, testing, prototyping, and perfecting a new product, agile innovators move quickly to come up with a minimally viable product that they can put into the marketplace for real-life testing, feedback, and adaptation. They use such digital techniques and tools as advanced simulations, 3D printing, and set-based design to accelerate the design process. They employ tight feedback loops to test, learn, and test again. The goal becomes not only product excellence but also continuous improvement based on customer usage and feedback.
Digital innovation presents traditional large organizations with multiple challenges. Technologies move fast; cycles times are short. The bets can be large and the uncertainties larger still. Traps must be avoided. One pitfall is attempting to apply digital technologies to existing processes instead of developing digital processes. Another is looking at digital technologies primarily as enablers of automation and greater speed—which they certainly are—but missing the chance to marry these technologies with human capabilities in order to create new ways of working.
While there are no roadmaps for digitizing innovation, there are plenty of models and laboratories that smart companies can use to test new ideas before committing to development. Digital natives usually test multiple ways of doing things, especially with respect to the collection and use of data. Companies can access laboratories in the form of model digital production facilities—such as those run by some universities and BCG’s own Innovation Center for Operations—which can be customized to illustrate the impact of assorted digital technologies and processes in various manufacturing, process, and production environments.
Companies can also make their own bets through M&A, partnerships, joint ventures, and participation in the digital ecosystems of organizations that spring up around emerging technologies. As we observed in our last report, more and more big companies are setting up their own digital venture capital funds, incubators, and accelerators to further their own experimentation. And big companies with an interest in potentially proprietary technologies are considering a variety of arrangements with so-called deep-tech startups, which are often more than happy to have a big corporate partner. (See “What Deep-Tech Startups Want from Corporate Partners” and “A Framework for Deep-Tech Collaboration,” BCG articles, April 2017.)
Whichever path they take, as they develop their digital innovations strategies and realign their organizations to function in a digital environment, companies will also need to digitize their innovation operations and processes. Those that try to produce digital initiatives with traditional approaches will soon find themselves mired in the old ways of doing things and frustrated over their inability to put digital technologies and ways of working into full operation.