It is no secret data science promises to make many aspects of supply chain management and procurement more efficient.
Much has been written about how data science techniques can improve automation, increase freight usages, track vehicles more accurately, enhance customer interactions and visibility, and help responses to external factors such as weather. But these improvements are only fully realised if businesses are aware of, and avoid, some very common problems that hamper the use of data science in procurement.
The first major problem businesses encounter is recognising the amount of data that is available does not necessarily equate to its effective use. Data science is part analysis, part artistry and consequently it needs to be grounded in a solid data infrastructure that makes use of the ‘right’ data sets – a process that is not necessarily intuitive.
Multiple sources of data from inside and outside a business need to be collated and crucially, updated as close to real time as possible. Without the full picture, data science cannot provide the most useful insights and advice to a business.
Not having the ‘right’ data is often a symptom of the second major problem, ownership and oversight of data science. Too often data science can be compartmentalised into one section of a business, for example, procurement. When really, a virtuous circle can develop where information and requests flow in from throughout a business, spurring better analysis and insights, which in turn prompts better understanding of the capability of data science within the business, encouraging more information sharing and ultimately better decision making. For example, efficiencies made in procurement could apply to marketing.
Similarly, knowledge or data held within the marketing or customer service departments could help inform the analysis of the procurement process. To get the best use of data science, the data science team or service provider should have visibility and buy-in from the whole business.
Finally, it is essential businesses realise data science is a much more powerful tool than traditional data analysis. Unlike most analysts who ask their data ‘what’, data scientists are programmed to ask ‘why’. By understanding ‘why’, data can reveal underlying, disparate, and often counter-intuitive, trends and factors that affect how a business functions.
This can lead to innovative solutions for inventory management by ensuring that influences on stock forecasting such as weather and customer profiling are incorporated. Transport and logistics can be made more efficient by planning optimal routes, improving notifications based on geo-location and the incorporation of complex delivery preferences. Put simply, you cannot get this level of insight from a data analyst.
Data science can be a silver bullet that improves how organisations relate to and serve their customers and employees. But it is not enough to employ data scientists or hire a data science agency. To be successful, data science requires access to the right information, support from the entire business and its full capability needs to be recognised. This is as true for using data science in procurement as it is for marketing, customer service, HR or sales.
☛ Mike Weston is CEO of data science consultancy Profusion