• inside

Improving Your Planning Process Productivity

Attitudes towards improving supply chain planning productivity are changing. Firms are beginning to ask why planners
need to spend so much time nursemaiding their planning system. Large enterprises ask, “What’s wrong with my
process that I need armies of planners?” Growing mid-market companies ask, “Why do I need to keep adding so much
overhead?”

Firms are asking, worse yet, if all this non-value added effort is preventing them from reaching higher levels of planning
maturity. Is their planning inefficiency preventing them from achieving advanced supply chain competitiveness?
Whether you call it “automated planning”, “lights out planning”, “low touch or no touch planning” or even “driverless
supply chain planning”, reducing human intervention is gaining currency. We call it “Powerfully Simple”, meaning that
the software is highly intelligent and built to run in a more autonomous manner, yet simple for the user. Being smarter, it
greatly reduces manual effort—dramatically improving planning efficiency and lowering the cost of ownership.

Here are some examples:

  • Lennox Residential has achieved 99.7% no touch, computer-controlled automation in their planning and replenishment. At Lennox, 997 out of 1000 planning decisions have been automated to the point where there is no manual intervention.
  • At Costa Express, just one planner now handles the planning for 5600 points of sale. Unmanned coffee stations transmit POS data every 15 minutes to help a highly autonomous planning system forecast demand, optimize inventory, and generate replenishment proposals for distribution and procurement.
  • Internet retailer Wayfair reported a 6 to 1 reduction in planning workload by moving to low touch forecasting and supply chain planning.
  • At pharmaceutical firm Cipla Medpro, the automated statistical forecasting system is now consistently up to 20% more accurate than Cipla’s own market intelligence. Says the supply chain executive director, “We’re now at the point where we can confidently switch off our manual overrides and put complete trust in the forecasts.”
  • In the UK, the National Health Service (NHS) is now using telemetry to send inventory and demand signals to a highly automated planning system to manage the nation’s blood supply chain.

Oceans of Demand Data

What going on? As products proliferate, demand complexity increases, and customers expect faster service, planners
are overwhelmed trying to create accurate forecasts and optimize inventory and replenish stock. Swimming in oceans of
data, planners can no longer manually incorporate into forecasts the data available to them—market trends, seasonality
patterns, promotions, new product launches, projected life cycles, POS data, or even social media data that signals
customer sentiment.

Automating the bulk of this far-reaching forecast process frees planners to take a low-touch; “advisory” approach, occasionally fine-tuning these baseline forecasts with their supply chain domain expertise and knowledge of business upside and risk. Planners are still needed for more complex and particularly more chaotic problems, but simple, mundane and routine activities should be just that – routine – and therefore prime for automation.

Powerful statistical engines and machine learning can crunch quantities of data behind the scenes, adapting to plan and optimizing heterogeneous demand (both fast-moving and long-tail goods). Inventory optimization automatically translates customer service policies into the right inventory mix for profitable response (See diagram below).

loop

We call this moving planners from “in the loop” to “on the loop”. We didn't make this phrase up. It came from the world of semi-autonomous drones, where operators can leave much of the detailed decision making to the drone itself, but intervene with higher-level decisions.

Enabling Business Transformation

What’s also exciting about a low touch, highly automated approach to supply chain planning is that it not only improves productivity, but it can also potentially be transformative. Let’s take another look at how some of the companies mentioned above were able to transform their businesses:

– Costa Express used POS data, telemetry and rapid re-planning to enable an entirely new approach to logistics and replenishment. They now see themselves as redefining the ability to deliver premium barista-quality coffee in venues that were not previously practical.

– Now that nearly all planning decisions at Lennox Residential have been automated, they are starting to redefine their customer service around new business models. You can’t predict when an air conditioner is going to break down, but with Internet of Things (IoT) data streaming from the AC unit, Lennox will be able to diagnose impending failure points and contact the customer to arrange a service visit, enhancing the customer relationship and balancing their workload.

– In the brutally competitive world of internet retailing, Wayfair is able to support a customer-centric multi-sourcing strategy to compete with Amazon’s enormous investment in distribution and fulfillment center automation. Wayfair leverages their advanced demand planning and inventory optimization to provide them the agility to predict and respond to constantly shifting demand and customer fulfillment needs.

– Cipla Medpro has developed the intelligence to proactively identify potential stock-outs up to four months in advance. This gives them enough time to respond proactively to critically important customer demands.

– The UK’s National Health Service has revolutionized the process of maintaining the United Kingdom’s blood supply, transitioning from a push approach to a demand-driven ‘pull’ model.

Sometimes the transformative benefit is the ability to support growth and address the challenge of an expanding product and SKU-Location portfolio. For example, at Wayfair it took 3 planners 100 hours a week to create a forecast and replenishment orders for the growing assortment of SKUs. They implemented software that slashed planning time from 100 hours per week to 15, allowing the team to forecast an exploding portfolio of nearly 1 million SKUs and plan replenishment for more than 40,000 SKUs.

“Our team was snowed under with forecast exceptions and overrides,” said the senior director of inventory planning and sourcing. Now, “we can spend more time looking at the genuine exception SKUs rather having to question everything.”

A Segmented Approach

Having a clear end-game vision in mind does not always require accomplishing it in a single leap. The key to understanding and accepting the need for automation and analytics can be achieved by looking at the problem at a granular level and matching types of activities to those that can be automated. The right automation in the right place is not a job killer. It’s a job changer.

For instance, cleansing data and calculating baseline demand is an ideal application for automation. It’s a great place to start automating the things the computer does well, and separating out the tasks where planners excel. For more sophisticated problems like forecasting promotions or new product introduction, advanced analytics like machine learning, can raise the game further. Machine learning adjusts the baseline demand forecast by identifying the effect of demand indicators. It can “decode” both structured and unstructured data streams, analyzing the variables and indicators and their complex interactions and patterns in an automated fashion to "self-learn" demand profiles.

MACHINE

This allows the planner to be business-oriented—able to add value through market intelligence and business smarts. At UK wholesaler RS Components, global supply chain planning boss Andrew Lewis was a believer in automating demand planning and replenishment. He said, “For some time now, I've wanted to be able to separate off the creation of the statistical forecast from the enrichment of it by the demand planners." “I want my demand planners to be ‘people people,’ with interpersonal skills, who can bring out from other [business] functions the sorts of things the system can’t probably know… making the eventual consensus forecast—the demand plan—absolutely as good as it can be."

Do we need to keep nursemaiding planning systems, employing armies of planners, and adding overhead? No. Can we achieve higher levels of maturity by freeing up time to think and execute our way to advanced supply chain competitiveness? Yes. Many companies are already well along on the journey.