Now that the holiday season is behind us, are you thinking about all the orders you could have filled if your forecast, procurement, and production plans had been more accurate? Or are you worried about the orders you didn’t fill as you now try to get rid of excess inventory? Did you delight or disappoint customers.
Whatever your end-of-year supply chain performance blues may be, you can avoid being in the same situation this time next year by making one resolution right now. Open yourself (and your company) up to the idea of digital demand management. With it, you can improve the accuracy and frequency of your forecasts and plans, synchronize supply to demand and better respond to any variability.
So, say it with me: I resolve to adopt digital demand management in 2018. I will use it to optimize my flow of goods by sensing and responding to demand throughout my supply network to the delight of my customers.
Throw out outdated plans & go digital
It is surprising how much businesses often put at stake by relying on rules of thumb, averages and isolated instinctive decision making. Forecasts and plans produced on a weekly basis, using stale data, quickly become obsolete when operating decisions are made on a daily or even shift-by-shift basis. As a result, planners and schedulers routinely abandon their weekly plans, and end up using a crude mix of spreadsheets and tribal knowledge to make million-dollar working capital and operating cost decisions usually with less than expected customer service.
In contrast, digital demand management gives workers timely and even real-time data based insights to support day-to-day or minute-by-minute decisions, so it’s never outdated. It enables you to assess risk, set inventory ranges with control limits, and collaborate with other functions to leverage advanced analytics to predict if, and when, a process flow is likely to move out of tolerance. And it can even make automatic adjustments to supply chain processes based on detected changes – building powerful insight and embedding new business rules along the way.
Digital demand management starts with identifying your supply-demand process structures, flows and critical control points to leverage a network of IoT devices (like sensors or monitoring tools) that live at every control point in your supply network. These IoT-based data sources then converge with other digital initiatives – like mobility, machine learning and business intelligence – to connect silos and provide real-time visibility into the flow of goods across your entire supply network.
Supply-chain modeling, simulation, and optimization tools, use this aggregated data to determine where processes can be improved, or where roadblocks slow down productivity. They allow you to visualize and create an optimal flow of goods through your supply chain to match your desired forecast and response performance.
Rethink your forecast
Rather than forecast an imperfect absolute single-number, digital demand management allows you to forecast a range of acceptable outcomes and risks that you and other functions can agree to and work within.
You start by simulating what the flow of goods through your supply chain would look like if you hit the high end of your forecast. This gap, based on median deviations between the high end and the optimal flow represents your inventory cost risk (i.e. the cost of safety stock and excess inventory). Then, you simulate what your flow would be if you only hit the low end of your forecast. The gap you identify here is your service cost risk (i.e. the cost of running out of stock).
With this information, during sales & operations planning (S&OP) meetings, management can agree and set upper and lower control limits based on how much inventory and service cost risk you’re willing to accept. If you start to fall outside the predetermined control limits, analytics kick in to alert your teams to either ramp up or down production.
Advanced analytics add an important layer to account for variability. For example, predictive analytics can tap into historical patterns and flow data to match the actual flow at the time to determine if, and when, a process is likely to move to an out-of-tolerance state. And preventative analytics can use a history of corrective actions to guide workers on a resolution to the issue – or even use prescriptive analytics to automatically make a change without human involvement.
Get on the leading edge
Companies on the innovative edge are already using digital demand management to great success. For example, Proctor & Gamble collects advanced point-of-sale analytics from its trading partners. The company then uses the information to forecast and reschedule production daily, responding to demand variability.
Amazon uses in-line inventory optimization algorithms within its order-management system to determine which distribution center offers the lowest total cost to fulfill an order. This allows the company to dynamically rebalance its network inventory based on orders as they come in.
If you’re not actively working today to join these companies by developing and piloting digital demand-management processes, it may be too late. Predictive and prescriptive analytics require time for new sources of historical data to accumulate. And if you don’t resolve to start building the data foundation required for these machine-learning based applications in 2018, you may find it impossible to catch up to the competitors that did make that resolution.