Friday 15th Dec 2017 - Logistics Manager

Dictated by data

A few years ago, lumbering data warehouses with nightly number crunching were the rarefied preserve of computer science graduates. Today it’s rather different – data is the vital business driver.

While governments and spin doctors may agree with Josef Goebbels, who said: ‘He who runs the information runs the show,’ it’s a maxim that seems just as appropriate to the supply chain in our information society.

Back in the 1980s, retailers would say, perhaps a little cynically, that the best use of electronic point of sale data was to prop open the office door. Remember those reams of green and white listing paper? Today EPOS data drives not only retail replenishment, staff scheduling and performance management but production, raw materials sourcing and much more.

As Niall O’Doherty, global manager manufacturing/supply chain industry marketing for Teradata, put it at the recent Teradata Universe conference in Warsaw: ‘EPOS data now needs to drive processes well beyond retailing and CPG. Its impact is felt way back along the supply chain, affecting areas such as service parts and oil – and it also triggers logistics activity.’

EPOS data tells you more than what is selling, it provides that touch point on consumer activity that those further along the supply pipeline are eager to access. It is also driving demand-based forecasting techniques such as flowcasting, directing freight movements and global sourcing.

Selling direct

The internet and direct sales also deliver that vital consumer touch point so it’s not surprising that more manufacturers are opting to sell direct, albeit often discreetly and under pseudonyms to avoid upsetting their retail customers.

Of course, Dell has no such problems and its entire business model is built around rapid reaction to buying trends. ‘Dell’s model depends on real time data,’ says Nick Hartery, senior vice president manufacturing, Dell. ‘We own our value chain which is shorter and more
integrated than in many other models.’

At any time, Dell can see what customers are ordering and in which geography so that it can react instantly to switch supplies or amend website offers to reflect stock availability. Component orders are placed every couple of hours and product can be assembled and shipped the same day. The backlog of orders to be processed is never more than two days.

‘We acquire materials at the last minute before the product is assembled,’ says Hartery. ‘We have an inbound logistics centre with eight to 10 days’ supply, but that’s not owned by Dell.’

Basic transactional data is not confined to the high street or internet. Inputs from call centres and service engineers are increasingly being used throughout industry in predictive analytics to improve spare parts management or even product design. ‘Early warnings
are vital,’ says O’Doherty. ‘If data about product performance comes back quickly then manufacturers can react to that at a component level to correct faults, trigger recalls or change suppliers. Unlike forecasting, these sorts of predictive techniques are less concerned with accuracy and more involved with identifying trends early.’

As O’Doherty also points out, the automotive industry may like to sell extended warranties but if these result in a high level of claims they are obviously less profitable. Profit comes from an absence of failure rather than from fixing problems, so early insights into any weakness, identified by predictive techniques, are essential.

Collecting this vital raw consumer data may be easy for high street chains and online players but the further down the pipeline you go the harder it becomes. As, unconsciously echoing Goebbels, O’Doherty puts it: ‘Information delivers competitive advantage so
companies are reluctant to share data as it’s a key differentiator for their businesses. Manufacturers are pushing to achieve greater customer intimacy and to do that they have to be innovative as they can be far removed from those customers.’

As numerous CPFR initiatives have demonstrated, to succeed, data sharing has to deliver mutual benefit. In retailing that is usually measured as fewer stock-outs. For outsourced call centres asked to report patterns of consumer behaviour, there need to be standardised and automated systems for data entry.

DR. PENELOPE ODY IS A REGULAR COLUMNIST WITH SCS AND IS A RETAIL MARKET SPECIALIST