All News & Insights

Digital Innovation – Scaling for the Fast and Furious Future

To fully embrace Manufacturing 4.0, companies need to prioritize a data strategy that’s focused on change and how to handle it. By Sath Rao


C-suite executives are strategizing for both growth and resilience. They all agree that they need to run their organizations for today but also need to transform for the future. There is a need for steady-state incremental change while retooling for the transformative future. This creates, in essence, a two-speed need for innovation. Business model transformations will require a new way of looking at data as a competitive advantage, and the enormous amount of data being generated will require new approaches to drive innovation, upskilling, and accelerating the idea-to-execution cycle.

These new approaches won’t be created through big budgets alone. It is well documented that giving a blanket budget often results in the funding of science projects. We have a plethora of pilot purgatories that have no chance to scale. Historically, the learnings from these projects are often ignored, resulting in more of the same. More recently, compelling arguments have been made about challenging innovation leaders to test more ideas with the least amount of money. Most companies continue to fund too many ideas rather than quickly killing those that don’t work.1

Steven Johnson in his famous TED Talk,2 “Where Good Ideas Come From” outlines how innovation is not always a sudden Eureka! moment, but a slow hunch that fades into view over time. This process of connecting the dots often requires collaboration with others, and chance always favors the connected minds. As we look at enabling change, Johnson says the innovation continuum can extend beyond doing the same thing, better — moving within the run cycle of business — to using digital technologies to fundamentally transform a business model with continuous change, i.e., the transformation cycle. All along this continuum, the common thread is data. A digital transformation cannot be charted without a robust data strategy as the underpinning of the journey.

Manufacturing transformation or industrial transformation is viewed as keeping in step with the larger enterprise-wide digital transformation strategy. It can run parallel paths with independent innovation opportunities that are often vital and central to the core digital transformation tenet. The big challenge is in understanding the anatomy of change, which depends on several exogenous industry factors including threat from competition and degree of industry change, as well as several endogenous factors like organizational culture, decision-making philosophy, and leadership. Ultimately, several of these disconnects lead to the famous IT-OT divide, which is a serious impediment to the rate of change and the push-pull unison that IT and OT must demonstrate for successful transformation.

So, is there a data-imperative for Manufacturing 4.0?


Manufacturing 4.0 requires a seamless data flow and, more importantly, a robust data strategy.

Key Components to a Smart Factory  

What is needed for a factory or systems to be in step with Manufacturing 4.0? In his book, Data Strategy,3 Bernard Marr stated that a “smart factory” must include the following four features:

  • Interoperability, with machines, devices, sensors, people, and systems interconnected to seamlessly exchange data in real time or relevant time.
  • Information transparency, in which cyberphysical systems contextualize what’s happening on the shop floor by creating a digital twin of the manufacturing process.
  • Technical assistance via human-machine collaboration, through which machines can support humans in making decisions and solving problems, as well as assist them with difficult tasks or in hazardous environments. Likewise, humans must be in a position to communicate and manage machines, especially when operating in dangerous situations.
  • Decentralized decision-making, where cyberphysical systems can make basic decisions autonomously to support routine tasks, and data can easily move down or up — from the edge to the cloud, for example — to support decisions at multiple levels.

Clearly, for this vision of the smart factory, a data digital thread is a sine qua non! The IDC Industry Spotlight on Roadmap for IT/OT Convergence predicted that by 2021, 80% of all industrial companies will have merged operational data streams and enterprise data streams to support broader and more rapid operational innovation.

Businesses are actively embracing digital transformation and the key tenets of Manufacturing 4.0. In its 23rd annual global CEO survey,4 PwC found that 77% of global CEOs (83% in the U.S.) plan operational efficiencies to help drive growth. Additionally, the report concluded that CEOs in the U.S. are “doubling down” on their commitment to pursuing operational efficiencies in 2020, an increase from surveys in recent years. PwC attributes this uptick to the emergence of “intelligent automation and cloud-enabled tech services” that are enabling these CEOs to turn uncertainties (such as those from trade conflicts and supply-chain disruptions) into opportunities. These newly available technologies are providing the foundation for companies to build supply chains and sourcing strategies driven by data, automation, AI and ML.

Elements of a Robust Data Strategy 

What do the four requirements of a smart factory all have in common? Data. Not only is it the underlying theme, it’s the driving force behind achieving the next level of business transformation. Manufacturing 4.0 requires a seamless data flow and, more importantly, a robust data strategy.

Change is a complicated beast. Manufacturing companies need to empower both IT and OT to own a transformative vision for change, to help them understand what needs to change, and how to fund change. With a focus on a use-case roadmap that is part of the multiyear digital transformation journey, companies can achieve competitive advantage with change and the right data strategy.


A data strategy must answer crucial questions in four key areas, all of which work together.

  • Transformation. What is a transformative vision for change? How can a competitive differentiation include customer experience?
  • Ownership. Who is responsible for driving change? Who owns the data? What are the privacy and security measures?
  • Change and funding. Which specific processes and architecture will need to be affected to enable the transformation initiative? How do you fund change to sustain transformation?
  • Win with change. How will relationships between humans and machines change, and how can they best complement each other in making winning choices that support critical business outcomes? How can data-driven innovation be accelerated by upping the data skills and capability of the organization to learn?

According to Marr in Data Strategy, a good data strategy is one that focuses on business value — what your business wants to achieve and how data can help you achieve it — and not on whatever data is currently or potentially available. The ways in which data can help a business succeed are many, says Marr, but in essence it comes down to “using data to improve your decision making, using data to drive operational improvements, and treating data as an asset in itself.”

When manufacturers deeply understand each of these areas for their business and develop a thoughtfully planned data strategy based on this understanding, they can implement a robust and scalable foundation to support their journey to Manufacturing 4.0. Here I’ll explore the anatomy of each of these elements in more detail.

Anatomy of Change 

1. What needs change?
The first key component of the data strategy in digital transformation is understanding what prioritized aspects of change are needed across the entire manufacturing organization. It is important to consider the shorter-term run as well as the mid-to-longer term transformations needed as part of a multi-year roadmap.

Differentiating on sustainable competitive advantage that includes customer and employee experiences will also be increasingly important. In many ways, customer centricity is the next major wave of change. In the coming years, every manufacturer’s data strategy will need to address how to gain in-depth visibility into customer product and service use patterns, preferences, and other criteria required to manage the customer experience.


To deliver the outcomes organizations require, industrial transformation must be fully aligned with business requirements and processes. Although manufacturing teams can contribute the most extensive knowledge in their domains of OT experience, they must look beyond what’s happening on the shop floor and synchronize it with where the business needs to go. Likewise, there are several areas where IT can help on the shop floor with best practices around security, for example.

2. Who owns change?
The second major component of a robust data strategy is to understand who owns change. In most cases, motivating an organization for change is a top-down initiative, driven by the CEO, CIO or Chief Data Officer (CDO). However, the ownership of change is not always simple to define. Roles are continually evolving, and responsibilities are often shared. Whatever their role, leaders propelling transformation require an effective data strategy — having the data is not enough. They need to consider what types of data can enable transformation and how they can best use the data to achieve their goals.

Understanding who owns change is crucial because it affects the span of the organization’s data strategy. Individual responsibility within areas of expertise are important, but a data strategy that will serve the needs of the entire company must encompass all aspects of data, including governance, security, and privacy. For example, a plant manager will be focused on operational change, whereas a CDO is more likely to fully understand the business as well as technical priorities that must be considered to benefit the entire company.

For example, in a customer-centric environment, the type of data being generated is everchanging, as organizations accumulate vast volumes of structured and unstructured data. If the data includes video or other sensitive personal information, privacy often becomes a priority.

The CDO position in the organizational hierarchy also plays a critical role in the ability to catalyze change and a coherent data strategy. Managing data governance in an inherently non-digitally native organization has challenges. There is often a need for close cooperation with the CIO’s office to develop a unified view around data quality, data governance, master data management, and developing a smart-analytics architecture.


Understanding the anatomy of change helps businesses evolve to a new paradigm of continuous transformation.

According to a recent Gartner study,5 fewer than half of CDOs said that they reported to the CEO. Their role is that of a change agent and communicator of the future vision of the organization. The enabling next steps they can promote should include upskilling the organization and enabling decentralized data-driven innovation.

In “The Case of the Missing Insights,”6 I examined how manufacturers can use data to drive value using artificial intelligence (AI) and machine learning (ML) technologies. The article discussed the importance of prioritizing the right problems in the run-transform cycle and how focusing on business value will help firms derive benefit from the power of these technologies. The CDO (in conjunction with the CIO) often has to be the enabler of this vision, working with diverse stakeholders to help nurture the inquisitive organization that asks the right questions and identifies the right problems to be solved using the right solution approach. Ultimately, the ownership of change has to percolate to the empowered individuals who are now supercharged with the vision of transformation.

3. How do you fund change and win with change?
A digital transformation initiative must carefully consider not only what needs to change but how. For example, consider the case of a leading automotive manufacturer seeking operational improvements, which is the “run” cycle of business. The use-case prioritization could be based on operational savings, the return on investment, or a reduction in quality issues and scrap. For a more comprehensive data management strategy focused on the “transform” cycle — that is, recognizing that the future of mobility lies in electric and autonomous vehicles, both of which rely heavily on data — the organization will need a completely different view on funding.

Change is highly complex and, without a clear roadmap that incorporates the means to funding change, it is all but a pipe dream. Manufacturers need to empower their organizations to own a transformative vision for change, understand what needs to change, and determine how they should fund change. One way to start is by focusing on a specific set of use cases that are part of a long-term digital transformation journey — all driving toward winning competitive advantage with change.

4. Democratized decision-making accelerates digital innovation
Decision-making is the final element of a comprehensive data strategy and the one that best defines the relationship between people and technology. Enhancing decision-making is a fundamental objective of data strategy. One of the most compelling capabilities of a data analytic framework is its ability to help eliminate bias from human decisions. In their book, Prediction Machines, The Simple Economics of Artificial Intelligence,7 authors Ajay Agrawal, Joshua Gans and Avi Goldfarb articulate the fine line between bias and judgment.


Change is highly complex and, without a
clear roadmap that incorporates the means
to funding change, it is all but a pipe dream.

Bias is centered around a highly personal, sometimes unreasoned favor for or against a decision. In contrast, they say, judgment is a much more active process that is forward-looking and that fully considers the pros and cons of an outcome. Making a decision involves an element of prediction of a possible outcome, as well as an application of judgment to decide what needs to be done. When actual outcomes are examined in hindsight, they provide ongoing training to human beings as well as to AI models.

The authors note that machines are generally incapable of providing judgment. This is something that only humans can do, because they are uniquely able to express the relative rewards from taking different actions. As AI becomes more adept at accurate prediction, humans will gradually make fewer decisions involving both judgment and prediction. Instead, they will increasingly focus on making judgment calls. The authors predict that this shift will enable an interactive interface between machine prediction and human judgment, just as people currently run alternative queries as they work with databases or spreadsheets. The system will provide decision-makers with a set of possible outcomes and their associated risks.

Decentralized Decision-Making 

Understanding the anatomy of change helps businesses evolve to a new paradigm of continuous transformation. Until recently, the perception was that an individual who was higher up in an organization would have more visibility into relevant business metrics and therefore could make better decisions that would impact profitability. However, as noted above, the move toward Manufacturing 4.0 is increasingly driving the decentralization of decision-making. As data volumes increase and data analytics gets more sophisticated, manufacturers can provide the same level of insights, broken down into a variety of metrics, to decision points all along the process down to the shop floor.

This new paradigm is centered around humans, machines, and data working together, which in turn drives competitor advantage and creates additional value for the business. Digital transformation in manufacturing is rapidly changing the primary focus away from economies of scale to a focus on economies of learning. These economies of learning can help reduce or even eliminate human biases from decisions, leading to faster, better decisions across the enterprise.

Improved access to data also has substantial potential to confer new network effects. In the recent past, solution providers focused on creating platforms. This focus led organizations to believe that they would realize substantial advantages by simply adopting a new platform, but that model has not always succeeded in the manufacturing industry.  

Digital transformation is happening here and now, and data is the wellspring of the next wave of manufacturing.

The focus today is increasingly about asking how that level of decision-making can be provided to end users and whether they can impact customer experiences. These customer experiences have the potential to create a network effect because customers tend to value consistency in their experiences. This can help organizations achieve the customer-centricity they require to realize competitive differentiation.

In a recent article in the Harvard Business Review,8 Andrei Hagiu and Julian Wright compared the network effects of platforms with the potential network effects of data-enabled learning. On a platform, such as a social media platform, a network effect is achieved when the platform increases in value as more people use it, eventually reaching a critical mass that shuts out the competition. According to the authors, data-enabled learning can achieve similar results. When learning from one customer creates a better experience for other customers, and when the company can incorporate that learning into its product quickly enough to benefit its current users, consumers become invested in how many other people are adopting the product — creating a network effect similar to that of a platform.

Hagiu and Wright predict that, in the years ahead, “improving offerings with customer data will be a prerequisite for staying in the game.” Most often, however, that alone will not be enough to win that game. Instead, they predict that the businesses that will be the most powerful and valuable in the future will be those that take advantage of both traditional network effects and those enriched by data-enabled learning. And some businesses today are already exhibiting those traits, such as Apple’s App Store, Facebook’s social networks, and Alibaba’s and Amazon’s marketplaces.

Go Big With Data  

Digital transformation is happening here and now, and data is the wellspring of the next wave of manufacturing. The data imperative is about redefining data strategy at an organizational level.

While all four of the elements are essential to developing a robust data strategy, at a fundamental level, digital transformation in the manufacturing space is all about the anatomy of change.

To achieve the four fundamentals of the smart factory, organizations need a comprehensive and robust data strategy that is fully aligned to business imperatives. If you want to get to Manufacturing 4.0, you need to think big about your data. The future belongs to the fast and furious learning organization!   M


3. Bernard Marr, Data Strategy: How to profit from a world of big data, analytics and the internet of things (London, Kogan Page Limited, 2017).
5. Rethinking the Role of a Chief Data Officer, Forbes Insights, May 22, 2019.
6. Sath Rao, “The Case of the Missing Insights,” Manufacturing Leadership Council, October 4, 2019.
7. Ajay Agrawal, Joshua Gans, Avi Goldfarb, Prediction Machines: The Simple Economics of Artificial Intelligence, (Boston, Harvard Business Review Press, 2018)
8. Andrei Hagiu and Julian Wright, “When Data Creates Competitive Advantage,” Harvard Business Review, January-February 2020.

View More