Think at London Business School
365体育投注Friday 13 March 2020
365体育投注Leaders who test new ideas robustly using scientific methods can then draw firm, valuable conclusions from them
By Julian Birkinshaw
As a professor at London Business School, I often have the privilege of working with amazing organisations that constantly push the boundaries of what is possible and make a positive impact in the communities they operate in.
One such organisation is Guy’s and St Thomas’ NHS Foundation Trust – the hospital is famous for many achievements, including being the site where Florence Nightingale founded her nursing school 160 years ago. She professionalised nursing and ushered in the era of hospital care as we know it.
While working at a British military hospital during the Crimean War, she used data to establish that poor hygiene was, at least in part, responsible for the high mortality rate among injured soldiers. She created compelling to illustrate her point, designs that are still used by statisticians and process control engineers to this day. Through data she was pivotal in .
Modern hospitals today collect more sophisticated data than they did in Nightingale’s time. From the moment a patient arrives to the moment they are discharged, every one of their interactions with a physician or a nurse, every diagnostic test and every procedure performed is recorded in the hospital information system.
365体育投注Beyond being a useful record of the care provided, one can imagine that such a rich dataset would have a huge number of operational uses – for financial planning, for identifying capacity bottlenecks and assessing whether interventions, such as employing more doctors during the weekend, actually work. With a few exceptions though, this data is not used as much as it could be.
“The future belongs to those who are able to lead and ‘speak data’”
In my experience, hospitals are not alone in this. Retailers, insurance companies, banks, utilities, transportation, pharmaceutical and consumer goods firms, to name a few, all have access to vast amounts of data about their customers, their suppliers, and their internal operations. And this data is underutilised. The main reason for this improvident state of affairs is the lack of data-analysis capability – the ability to use data to influence decisions.
How can this be? The answer is path dependency – these organisations are led by talented managers who have risen through the ranks because they understand their industry, have generated value for their stakeholders and demonstrated their ability to lead others; skills that typically do not involve the ability to analyse data.
365体育投注In the past, when data was not so readily available, this was less of an issue. But in an increasingly digitised world, where data is becoming cheaper to collect and more readily available, the future belongs to those who are able to lead and ‘speak data’.
365体育投注The language of data analysis involves both practical skills and cultural imperatives. On the practical side, a senior manager needs to understand what data science can do – the value of data that can be found in:
Just as importantly, a senior manager needs to understand the limitations of data science. They need to know when a problem is too complex, too difficult to analyse or when the data available is ill suited to the problem at hand.
“Like any language, data analysis can be learned”
On the cultural side, speaking data requires the organisation to create an expectation that decisions have to be backed with evidence. A company that comes to mind that has achieved this is Amazon where data-driven analysis is an integral part of every process, from optimising the website to optimising the warehouse. Initially, the benefits of such an approach may be small but as Amazon has demonstrated, over time a lot of smart continuous improvements can generate a big competitive advantage.
365体育投注Like any language, data analysis can be learned. Again, like any language, it may be easier to learn when you’re young. But in teaching data science at the School for 12 years now I’ve observed that with a bit of effort, the right motivation, and a little guidance anyone can pick it up. You don’t need to become a data-science expert.
Instead, the expectation is that you will continue to be an expert in your industry, but with a healthy appreciation of what data science can do. You will be able to do simple analysis for yourself, but perhaps more importantly, become a savvier consumer of the data-analysis work of others and a more effective manager of more junior colleagues who are tasked with the analysis.
To help more senior executives achieve this, my colleague Tolga Tezcan and I have created Data Science for Business Intelligence. As part of this course we ask participants to bring a data set from their context and to work on it in class. Thus, the data science methods we discuss in class are no longer abstract but can be put to good use immediately. This not only benefits the participants but also teaches me new ways in which data science can generate value; or as Florence Nightingale put it, an “administrator could only be successful if [s]he were guided by statistical knowledge”.
Nicos Savva is Professor of Management Science and Operations at London Business School.