Big Data and the Innovation Cycle
by Hau L. Lee
Production and Operations Management
Vol. 27, No. 9, September 2018, pp. 1642–1646
DOI 10.1111/poms.12845
Big data and the related methodologies could have great impacts on operations and supply chain management. Such impacts could be realized through innovations leveraging big data. The innovations can be described as first improving existing processes in operations through better tools and methods; second expanding the value propositions through expansive usage or incorporating data not available in the past; and third allowing companies to create new processes or business models to serve customers in new ways. This study describes this framework of the innovation cycle.
Key words: big data; innovation cycle; supply chain re-engineering; business model innovations
- Introduction
Big data and the related methodologies to make use of it: data analytics and machine learning have been viewed as digital technologies that could revolution- alize operations and supply chain management in business and society at large. In the 2016 survey of over 1000 chief supply chain officers or similar senior executives, the SCM World found big data analytics at the top of the list of what these executives viewed as most disruptive to their supply chains.
We have to be realistic and recognize that the use of big data and the associated development of tools to make use of it is a journey. This journey is a cycle that techno- logical innovations often have to go through, and at every stage of the cycle, there are values and benefits, as well as investments that we have to make in order to unleash the power and values.
- The 3-S Model of the Innovation Cycle
new technologies often evolve in three stages.
The first one, which I called “Substitution,” is one when the new technology is used in place of an existing one, to conduct a busi- ness activity.
The second one, which I called “Scale,” is one when more items and more activities are used with the technology more frequently and extensively.
The third is the “Structural Transformation” stage, when a new set of re-engineered activities can emerge with the new technology.
- The Substitution Stage of Big Data
The availability of big data can immediately allow new methods or processes to be developed to substitute existing ones for specific business activities. An obvious one is forecasting. Much deeper data analytics can now be used to replace previous forecasting methods, making full use of the availability of data. Such data were previously not easily accessible.
- The Scale Stage of Big Data
Back in 2011, Gartner has identified the three Vs of big data: Volume, Velocity, and Variety (Sicular 2013). Rozados and Tjahjono (2014) gave a detailed account of the types of data that constituted the 3Vs. There, they described that most of the current usage of big data had been centered on core transactional data such as simple transactions, demand forecasts, and logistics activities.
Manenti (2017) gave the example of Transvoyant, which made use of one trillion events each day from sensors, satellites, radar, video cameras, and smart phones, coupled with machine learning, to produce highly accurate estimates of shipment arrival times. Such accurate estimates can help both shippers and shipping companies to be proactive with their opera- tions, instead of being caught by surprise with either early or late arrivals of shipments. Similarly, Manenti (2017) reported the IBM Watson Supply Chain that used external data such as social media, newsfeeds, weather forecasts, and historical data to track and pre- dict disruption and supplier evaluations.
- The Structural Transformation Stage of Big Data
Ultimately, companies can make use of big data to re- engineer the business processes, leading to different paths of creating new products and serving cus- tomers, and eventually, potentially creating new busi- ness models.
Product designers will leverage data on fabric types, plastics, sensors, and most importantly, connectivity with customers. Real and direct customer needs are used to generate new products, identify winners, and then work with partners to produce the winners at scale.
Making use of data on items on web pages browsed by e-commerce shoppers, Sentient Technologies also created machine learning algorithms to do visual cor- relations of items, and delivered purchasing recom- mendations (Manenti 2017). Again, a new product generation process has been developed.
I believe there will be many more opportunities for big data to make similar disruptions to the
- Concluding Remarks
it is often the case that a new innovation requires many small-scale pilots to allow early users to gain familiar- ity as well as confidence, ascertaining the values that one can gain from the innovation. Such early usage had often been based on one particular business activ- ity or one process of the supply chain.