Dharmendra Patwardhan, Executive Vice President of Business Services at Capgemini
Already established as one of the most talked about back-end business solutions of recent years, Intelligent Process Automation is set to remain on top of companies’ agendas throughout 2020. But while many continue to herald the technology as the key to unlocking all their problems, the reality is a little more complicated.
Because while intelligent process automation can solve problems – and also create opportunities – in the supply chain, it cannot do so unguided. In order to reap the benefits of automation, businesses need to point the technology at what they want it to fix – and therein lies the much of the disconnect between the promise of intelligent process automation, and the reality.
Intelligent process automation is rooted in practicality. It’s not about technology looking for an application: it’s about looking first, and in detail, at processes, addressing issues, and streamlining tasks, before automation and robotisation technology is brought into the picture. When this is applied to the supply chain, a number of opportunities present themselves.
So, the big question is: which processes should you be considering for intelligent process automation? This article will run you through four key areas where it can make a quick and vital impact to the supply chain.
Picture a major soft drinks operation. It consists of a company that owns the brand and makes the concentrate, and a number of other companies, acting in the manner of a consortium, who bottle and distribute the end product. Now imagine the consortium decides to run a price promotion of 97p per can, against the usual price of a pound. The potential problem with this promotion is that individual bottlers have different systems, and some of them may not recognise a single can as an entity – in which case, the promotion will either be applied inaccurately, or it won’t run at all.
Typically, standard automation could be introduced to handle the discrepancies created by the thousands of three-pence claims from retailers. But by implementing intelligent automation, the problem would be approached differently. It would create a consistent platform, not allowing a unit price to be acted upon unless it resides in the system, which eliminates the need to have a team correcting data or handling claims.
In short, the difference between the two approaches is that standard automation addresses the problem after it’s happened, whereas intelligent automation looks for areas of inefficiency, and addresses them up front.
This ability to address problems in advance is also key to our second area of potential benefit – order validation.
At the moment, major organisations traditionally process orders in one of two ways: either manually, by responding to emails or PDFs; or via Electronic Data Interchange (EDI), which is a form of automation itself, from their customers.
Unfortunately, however, EDI’s automation is limited – individual products have ID codes, and the code assigned to an item by the customer may not be the same as that assigned by the manufacturer or supplier, so there is room for error. All EDI is doing is pushing data indiscriminately through the system.
One way we’ve employed intelligent process automation at Capgemini to address this issue, is by deliberately breaking the natural flow of EDI and passing the order information through an order validation engine. This creates a common and consistent data set – as before, which prevents a problem getting into the system or process up front, rather than having to deal with the issues it causes later downstream.
Intelligent process automation doesn’t just make existing processes more efficient – it also opens up supply chain opportunities that weren’t available previously, such as in digital demand planning.
Typically demand planning relies on sales history – when you know how well a product has sold previously you can make predictions about future demand. But when you’re bringing a new product to market, there is by definition no sales history, which make it very difficult to make forecasts.
Again, this is where intelligent process automation comes in. By bringing together statistical models and machine learning tools, an organisation can analyse products that don’t have concrete histories. After all, the majority of new products aren’t completely new territory – they are iterations of, or extensions to, other stock keeping units (SKU) and usually joining a pre-existing product family. By extrapolating data from similar, relevant SKUs, and leveraging data to compare the forecast and the actual sales history, a planner can make a data-based informed decision of the sales forecast for the new product. This gives planners a statistical frame of reference that wasn’t available to them before.
Sticking to the theme of planning, one final area (at least for this article) in which intelligent process automation comes into its own is promotion planning. This is an area of special importance in consumer goods and over-the-counter pharmaceuticals, because promotions account for a significant proportion of overall revenue.
The challenge that companies in these markets face is that they tend not to keep libraries of past promotions (the nature of the offers, along with their expected and actual effects on sales). And because of this, they can’t be confident of the expected uptick on future planned incentives.
However, by revisiting historical sales data and applying machine learning techniques to gauge forecasts against reality, companies will be able to create a library that didn’t exist before. Armed with this, organisations have real data to base their future promotion decisions on.
With company margins continuing to tighten, businesses will be looking for ways to make processes more efficient and less time-consuming. The supply chain is an ideal candidate for this streamlining effort, with intelligent process automation offering a number of ways to help simplify procedures that are no longer fit for purpose – but only if you first hunt down the inefficiencies yourself.
When done successfully, intelligent process automation is about the application of digital transformation principles to specific individual scenarios – enabling us, in the supply chain and in other areas of the enterprise, not just to solve perennial problems, but to create exciting opportunities for innovation and growth.