In search of the perfect order: How Decision Intelligence can help master order fulfilment

Decision Intelligence

Back at the start of the millennium, I was fortunate enough to find myself part of a team empowered to redesign a global CPG company from the ground up, in what was to become the world’s largest SAP implementation. Responsible for the customer-to-cash processes, one observation I made was that across the different business units the determination of ‘who got what stock’ was consistent only in its inconsistency.

By Sean Culey

One variable that was consistent was the fact that customer priorities were made by individuals, not data. Decisions were taken in isolation without consideration to the total demand and supply situation and with no understanding of the long-term impact, and often the customer service team only found out there had been an issue when the customer rang them to complain.

To resolve these issues required the implementation of a new set of tools and processes to bring structure and discipline into the demand fulfilment process and minimize knee-jerk production changes and expensive order expedites. The aim was to highlight significant demand deviations before they became an inventory issue, not after. Tools such as ATP, product allocation, backorder rescheduling and product substitution were used, a daily schedule of order analysis and handover points was agreed, a series of order pipeline and ATP failure reports produced, and a weekly ‘demand/supply review’ meeting was established. We called this new suite of processes ‘demand control’, because although we couldn’t control demand, we could control our response to it.

Twenty years later and the need for these types of processes have become well recognized, only rather than ‘demand control’ they have names such as ‘sense-and-respond’ or S&OE. However, they still require a series of integrated meetings where human employees come together to analyze reports and spreadsheets, make decisions, and manually update the system.

Unfortunately, the business environment is not the same as it was back in 2002. The pace of change in our online world, combined with demand volatility and supply uncertainty, means that a formal meeting process underpinned by analysis of reports and spreadsheets is likely to take all day, not 30 minutes. Monthly and weekly review processes are too slow when the market operates 24/7, and sales history is a poor predictor when the future looks nothing like the past.

No longer can the supply chain wait for humans to gather to reach consensus. Rather than relying on people to make decisions using alerts and warnings from transactional processing systems, we now need machines to provide real-time recommendations and act on our behalf. Deciding whether a customer order can be fulfilled requires control-tower style visibility of the impact of this decision horizontally across the end-to-end supply chain, and vertically in relation to plans and profitability.

The order fulfilment decision making process is complicated: Do we reallocate stock from other orders? Delay the order? Expedite it when stock is available? Split the delivery? Deliver from a different location? Substitute products?

Answering these questions requires insight into product availability and production and distribution capability:
● When can additional products be made?
● What is the cost of making more stock and the impact on our production schedules?
● Can we do it without incurring overtime costs?
● What is the cost of moving stock from other locations?
● What is the carbon impact of these actions?

The impact on multiple variables needs to be understood and considered to address these challenges:
● Is it better to delay, part ship or expedite?
● What is the impact on cost if we need to split picking and shipment?
● What penalties will we incur if we are late or short?
● What is the impact on order profitability if we expedite the order or ship from an alternate location?

The ability to instantly analyze these multivariate factors, simulate these scenarios, and answer these questions instantly across thousands of order items is beyond human capability. Luckily, we now have a new form of enterprise-wide, AI-enabled software that can do just this.

This new capability is called Decision Intelligence, and it is powering a new form of cognitive control tower that leading companies are using to make the best order fulfilment decision possible. Instead of analyzing reports and making system updates, now systems such as Aera augment and automate human decision making, providing real-time recommendations and system updates to ensure that customers get what they ordered, when they ordered it, at the optimum profitability.

The perfect order delivered in the optimal way.

To see what this capability looks like, join the Webinar Wednesday with Declan Supple and Amine Benmesbah of Aera Technology on May 4, 2022. To see how Decision Intelligence can be applied to other areas across the business, the Future Now webinar series covers everything from sustainability to promotion planning.