Organisations, like natural organisms, experience evolution. And each new mutation – from the assembly line to the transnational corporation to today’s “flat” structure involving communities and crowdsourcing – has been traditionally the subject of rigorous study by academics. I believe that both researchers and practitioners would benefit from a new approach. The time may be right for academics to play an active role in organisation design innovation – just as university laboratories have long produced breakthroughs in computer science and biotech.
Outside consultants hired by companies to revamp their organisations rely heavily on practices that have worked before. “We did this at company X, and look how well they are doing” is the typical pitch. This methodology, which I will call “best practice transfer”, has two potential problems. First, correlation is not causation. It is easy to be fooled by randomness into thinking that a particular structure or practice drives success, when in fact it does not (or it may have even harmed performance). Placebo effects – where thinking that something will work can be self-fulfilling – may also occur. Second, organisations are all different; even something that worked well elsewhere may not work in your firm.
Techniques from the Valley
A suite of new methodologies based on digitalisation are available today that enable us to do better. Using historical “Big Data” from within an organisation – such as individual career trajectories, group dynamics and social networks – we can apply machine learning algorithms to help build predictive models. With computational agent-based models, we can simulate, cheaply, the impact of a proposed organisational change on outcomes without having to directly try it in expensive and risky direct applications. Finally, we can use randomised experiments (e.g. A/B testing, scenarios, field experiments) to test whether a practice will really have the intended effects in our organisation. Computational power and innovation in digital data capture are the platform technologies on which most of these methods operate.
With such techniques, organisational designers could use rapid prototyping and experimentation, techniques for which Silicon Valley has become famous, to test-drive various ways of organising the firm’s resources. This next phase of organisation design – I call it “Org 2.0” – can be a quantum leap beyond the older approaches of designing the “boxes and arrows” of org charts.
But the skills needed for Org 2.0 are not widespread in the corporate world today, or even most management consultancies. In fact, the HR functions of most large organisations (with a few exceptions in the tech sector) are heavily skewed towards the poets, not the quants.
The Org 2.0 elective at INSEAD
INSEAD is among the world’s largest contributors to the management consultancy global talent pool. It seemed a natural testing ground for my ideas.
Earlier this year, I introduced a new mini-elective to prepare INSEAD MBA students for Org 2.0 (and that’s what the elective was called). As far as I know, no course like it has ever been tried at INSEAD, or at any of its peer schools. I teamed up with a senior practitioner, Dr. Andreas Raharso (formerly of Hay Group), to design and deliver this course on both the Asia and Europe Campuses. PhD student Julien Clement was the main Teaching Assistant and a major contributor to the design, as well. We had an amazing set of guest speakers from Aon Hewitt, Oracle, and Mercer. McKinsey’s OrgLab, which I think is at the forefront in this space among the management consultancies, sent two speakers.
Arranged into groups of four to five, students were told to act as the “board of directors” of a hypothetical company, Chimera. The elective’s four exercises each tackled a discrete, realistic business problem:
• Will it work? Using A/B testing, (a.k.a. Randomised Controlled Trials) to know rather than guess which organisational redesigns will be effective.
• Where are the “hot spots”? Identifying problem areas that require organisation redesign, using graph (network) theory.
• Who’s a quitter? Forecasting employee retention and departure using machine learning algorithms.
• Will it spread? Using agent-based models to forecast the likelihood that an organisational change will gain wide adoption.
Each board received help from TAs – who acted effectively as “Chief Analytics Officer” – actually INSEAD PhD candidates who are well-versed in analytical techniques. The point of the course was not to transform MBA students into analytics experts, but to teach them how to collaborate with people who are. In essence, we trained them to ask the right questions to produce information that was both actionable and tailored to the desired outcome.
Before receiving the student evaluations, I didn’t know what to expect. Would our MBAs find it valuable to explore an arena so far outside the conventional purview of organisational development? Would they see the point to all the number crunching and playing with Python code?
Andreas, Julien and I were heartened to discover that the evaluations were extremely positive, and we have now expanded this elective to a full-length offering. Most encouraging to me was to see that students need not have come from a technical background to have benefited from the course. This was after all the point of our design. Of course, it helped that we launched it as a mini-elective, with a restricted class size. The expansion will have to be carefully managed. But the signs are promising that the Org 2.0 revolution just might take off!
Phanish Puranam is a Professor of Strategy, the Roland Berger Chaired Professor of Strategy and Organisation Design at INSEAD. He is the Academic Director of the PhD Programme.
Leave a Comment