The key challenge in organizational design isn’t big data, it’s people data
Article by Rupert Morrison
At the Organization Design and Big Data Conference in Paris, May 2013, speakers identified major impacts of big data, defined as the much greater volume, variety, velocity of data that we experience today, even compared with 5 years ago.
How does this impact business?
In marketing, big data allows more accurate customer segmentation. In healthcare, big data supports more targeted treatment. In recruitment, big data allows employers to screen more accurately. In the supply chain, big data reduces inventory wastage.
Does this actually change the way organizations are designed?
Jay Galbraith argued that big data does change organizations: it is forcing organizations to restructure themselves using big data as a new (fifth) structural principle. Some new functions will be the ‘big data’ operations which will distinguish themselves from the ‘rest’ of the business, just as the previous four structural principles separated an organization according business divisions, organizational functions, international expansion and customer segmentation.
Harry van Dorenmalen, Chairman of IBM Europe, also argued that big data was changing the ways businesses could work. For example within IBM, by collecting employee opinions through an open electronic town hall ‘jam’ session, and using its Watson software to process unstructured text, using billions of examples of text and processing algorithms.
At this stage, big data appeared to be an exciting technological innovation which affects many business models. Is it really a step change, however? And is big data really changing how we do organizational design?
I think that big data will have a big impact on how organizations work, but big data will not greatly affect the process of designing them. For the purpose of design, the complexity of the many-to-many links involved in people data are more important than the sheer size. I’d like to put forward some specific propositions:
- We will have to design different kinds of organizations to use big data.
- Our process for organizational design is unlikely to change because of big data
- The key challenge in organizational design isn’t big data; it’s people data
- Organizational design is not explained by economic theory
- Graph databases are changing the game of organizational design
- HR, Finance and Strategy need to evolve
We will design different kinds of organizations to use big data
Proposition 1 looks like an open and shut case. Organizations are already restructuring to deal with big data.
Jay Galbraith has cited P&G, which has created ‘control towers’ to maintain continuously updated control of its supply chain. He highlighted Amazon, which says it wants to be the world’s most customer-centric organization, partly by understanding its customers’ data in great depth. And he highlighted Nike, which created its Nike Digital Sports division in 2010, putting sensors in shoes, clothes and bands, and setting up virtual athletics communities.
Why should organizations restructure themselves to deal with big data?
Resource allocation becomes much more flexible in companies that can apply big data. With visibility of demand levels and supply volumes, they find it easier to move people, capital and other resources across sites, functions, roles and positions. For example, theme parks can reallocate labor quickly to busy areas; a supermarket business can respond rapidly to forecasts of changing weather conditions.
Today we find it normal that a supermarket chain should be able to understand the differential impact that the same weather conditions might have in different store locations and at different times on people’s behaviour; analyses as recent as Martha Starr-McCluer’s 2000 publication seem to be from a different age because the data sources and analytical techniques available at that time were relatively unsophisticated. Starr-McCluer, for example, had to work with average monthly temperature data across the whole of the US and compared that with average monthly sales data across 10 types of retail operations.
It’s clear that big data opens up possibilities that have never existed before to track the supply chain, customer-product interaction and the customer-company interaction. That will mean opportunities to deliver better services – which will mean different kinds of organizations. Interestingly, Galbraith argues that this will also generate some new tensions, and the most successful organizations will be the ones that manage the conflicts of direction and interest that will arise from having up to 5 different structural principles at work in the business.