By Trevor Miles
Velocity-based competition, shortened product lifecycles, more mass customization, increased demand variability, globalization and leaner supply chains have all altered demand management requirements in fundamental ways over recent years. And if you thought that was a tough environment, add the current economic crisis to the mix and you get a sobering picture of the challenges that demand management groups are facing on a daily basis as they try to ensure the right products are in the right place at the right time.
Traditionally, most companies have relied on historical data to predict demand. However, for a forecast to be truly accurate, the market and industry cycles must remain relatively consistent. That is not a realistic assumption anymore. Even before the current recession, companies struggled with demand volatility because of increasing competition, decreasing customer loyalty and constant new product introductions, among other reasons. But in today's grave economic climate, with major demand swings and declining spending overall, predicting demand based on historical data is now virtually impossible.
It is quite clear that historical demand is, at best, just one possible indicator of future demand. Therefore, statistical forecasts that extrapolate from historical trends should be seen as just one input to the demand planning process, not the entire process itself.
With demand so variable, companies have been forced to rethink their approach to demand planning and management and are now looking at how to:
- Get closer to real demand (understand point-of-sale, or POS, data) and ensure they have the full demand and supply picture.
- Do more frequent evaluations of forecast versus actual to determine variances.
- Develop a "bridging" process among internal and external partners to foster more collaborative planning and response.
- Ensure alignment of daily actions (inside the planning cycle) to corporate objectives.
At the end of the day, while demand planning and forecasting remains a core part of any business, the faster pace of change multiplies the problems that planning can't prevent. Leading companies need to excel not just at improving their forecast accuracy but also at their ability to effectively manage demand when the forecast is not accurate — which in many businesses is a significant portion of the time. Demand volatility guarantees that even the best demand and supply plans are inevitably out-of-sync.
So if you can't prevent change, you have to respond to it.
The financial impact to a company unable to respond to demand change can be crippling. Poor response can affect both the top line (i.e., inability to win new business, loss of customers to competitors) and the bottom line (i.e., negative impact on margins, write-off of excess and obsolete inventories) of a company.
As a result, new strategies and tools for not only better forecasting but also better responding to change have quickly become imperative for today's manufacturer.
21st Century Demand Management
Here's a checklist of key steps companies can take to address today's demand management challenges.
1. Get the Right Information
Create and manage forecasts at, or as close as possible to, individual stock-keeping units (SKUs).
- If your customer distributes your product to an end user, your sales data are not equal to actual demand — they have been modified by the distributor and may differ from real demand due to inventory, shortages, policies and promotions. In times of high volatility, this disconnect is magnified and the risks associated with the inaccuracy increases.
- The closer you can get to point-of-sale data, the better your demand forecasting.
- Ensure that you capture and consolidate forecast input from all the necessary disparate sources, i.e., forecasts from sales, customers and distributors across multiple distribution channels, as well as marketing forecasts and economic trends, plus actual demand history.
- Integrating this information, along with the unique considerations and insight from those involved, will help establish a consensus forecast, which will be your best starting point.