Similar questions are being asked by supply chain leaders around the world: Do we continue to invest in conventional processes that have proven effective but offer capped upside, or do we implement disruptive approaches that offer potential step-function advancement but also carry risks attendant with the unknown? In this article, we'll explore bimodal supply chain strategies that integrate elements of each approach.
What does "bimodal" mean?
Bimodal, as the name suggests, is the ability to flexibly operate in two supply chain "modes." Mode 1 leverages established, empirically proven, but often incremental approaches. Mode 2 explores new, potentially transformative, but often unproven approaches. Think of mode 1 as progression along a known, mostly linear function. Think of mode 2 as bending the improvement curve to access exponential gains.
The first mode centers on the delivery of well-established supply chain priorities like productivity, security, and reliability. Success in this mode requires a strong "microscope" perspective—near-term, detailed, precise. In mode 1, companies leverage core products, processes, and systems. An example of mode 1 is statistical process control—analyzing historical data to identify and reduce sources of variation. With more history comes larger sample sizes, which result in smaller standard deviation and more certain outcomes. Statistical process control is in the toolkit of any successful manufacturing organization. By systematically studying the past, we can consistently improve the present.
In the second mode, companies strive for step-function improvement that is rarely offered by the status quo, even if there is organized optimization of that status quo. This approach requires more of a "telescope" perspective—forward-looking, longer-term, less precise. Mode 2 involves exploring new markets, new processes, new technologies. The Internet of Things and machine learning, which leverage multiple data streams to self-adjust in real time, are examples of mode 2. As historical data is less available in mode 2, we have to look forward. Mode 2 can yield transformative advancements, although the statistical certainty of outcomes is lower.
Both is better than either
Rather than arguing for one mode over the other, bimodal recognizes the value of having both modes in your supply chain. The benefits of continuous operational improvement (mode 1) have been well documented over multiple decades. Likewise, the competitive advantages of leap-frogging conventional tools (mode 2) are similarly exciting.
Bimodal, however, shouldn't be thought of as a toggle, switching from one mode to the other. At the core of bimodal is the ability to integrate the two modes, complementing rather than compartmentalizing. Although modes 1 and 2 do require different perspectives, the real value of bimodal comes in connecting the modes. Data and digital technologies, further discussed below, provide one such link. Innovative organization designs provide another. Expanding one's view of "our" supply chain to include upstream partners and downstream customers is a third.
A challenge facing organizations looking to go bimodal is how to integrate the two seemingly disparate thought processes of modes 1 and 2. Many successful supply chains with whom we work—whether suppliers, partners or customers—translate deep process understanding into predictable advancements in important value-creation levers like cost reduction, capital efficiency, and service improvement. They leverage strong "microscope thinking" to mitigate risk while continuously improving, both hallmarks of mode 1. The "telescope thinking" needed in mode 2 often requires a different mindset—data automation and self-tuning algorithms replacing manual systems, end-to-end value-stream thinking replacing localized optimization, production-led team designs replacing traditional organizational structures. Mode 2's new approaches bring unknowns, while mode 1 is all about reducing uncertainty. This creates an apparent disconnect.
One common denominator between the two modes is data. Data is the language of continuous-improvement processes common to mode 1: Six Sigma, Lean and hoshin kanri (a method for ensuring that the strategic goals of a company drive progress and action at every level). Data also fuels the predictive analytics common to mode 2: cloud sharing, pattern recognition, and artificial intelligence. Digital technologies such as these that manage data establish a powerful link between the two modes, bringing together the individual strengths of each into a collectively more powerful union. Digitization, in both modes 1 and 2, builds trust and strengthens the links between partners by providing end-to-end value chain visibility that in turn benefits customers, suppliers, and communities.
Bimodal at 3M
3M is a $30 billion enterprise with more than 60,000 unique product offerings. We manufacture roughly 85 percent of everything we sell. With such a diverse supply chain, we rely on proven processes that have enabled us to continuously improve for more than a century. Like many reading this article, 3M is comfortable in mode 1. At the same time, 3M is also a science-based company. We see technology as a stabilizer rather than a risk. We're routinely ranked amongst the world's most innovative companies. So, while mode 2 is also new to us, it is generally familiar territory.
At 3M, we call our approach to bimodal "efficient growth." Mode 1 delivers the productivity gains that our customers and shareholders demand. Mode 2 builds on this foundation, fueling the growth necessary to ensure that our next century will be as successful as the last.
There are three key elements to our efficient growth strategy:
Together, these three strategies enable 3M to achieve higher levels of efficiency while also integrating new technologies necessary for innovation, growth, and customer loyalty. This helps us contend with the ever-increasing volatility of today while simultaneously exploring and innovating for tomorrow.
Follow this link for additional work from the analyst group Gartner on the bimodal model.