In this blog I will explore the need to ‘industrialise’ data science in order to realise tangible business benefits.
Getting value from skills investment
A key differentiator of businesses in the digital era is their ability to exploit data through the application of advanced analytics and data science. Organisations that recognise this will typically establish some form of data science capability. The starting point for this, unsurprisingly, is some form of data science capability and the recruitment of some data science expertise.
One of the issues organisations come across quite quickly, is how to drive value from this investment in expensive skills. Even if the Data Scientists manage to navigate the organisation and systems to deliver some insight, they typically lack the skills (not to mention the desire or interest) to convert that initial insight into tangible business benefit.
Industrialising the process
The analogy I see here is with the first industrial revolution. Whilst this revolution did depend on several scientific advances and the application of science to business, the key driver of value was actually the ability to industrialise the processes concerned. This continues to be a typical problem in manufacturing – the development of an initial model or prototype may demonstrate a great new product, but if it cannot be easily manufactured at a low enough price, then the invention is of no real value.
This is the exact same challenge that businesses face with data science today. Data Scientists are great at coming up with new insights, models or algorithms (‘products’), but they are typically produced in a one-off, hand-cranked, manner. To drive value from them, you need the skills to industrialise that process. Indeed, in the ideal world, the Data Scientist’s capability should be part of an end-to-end, factory style model that continually produces new, innovative products and then industrialises the production and monetisation of them.
The next blog in this series will look at the challenges associated with the industrialisation process.