Content area
Full Text
Outlines a methodology for forecasting pipefills of a new product . . . sales staff estimates are used as a forecast input... quantifies judgment with a Bayesian model ... sales surveys are used as a communication tool.
New products can be difficult to forecast. Here is a simple model from our forecasting toolbox which can help in this process. The objective is to forecast monthly pipefill quantities (initial orders from customers) of a new product. A marketing bonus from this methodology is a distribution penetration forecast. It has worked well for consumer packaged goods (CPG) such as batteries and consumer electronics. The tool uses a Bayesian approach for quantifying judgmental sales estimates of new product pipefills. Simply stated, it quantifies sales surveys using probabilities for success and timing. Typical demand variance for forecasts using this methodology has been lower than 30% during the first critical months, and less than 5% on hot products. However, strong management support and regular follow-up are the keys to success for this approach.
SURVEY THE SALES STAFF
This approach relies on sales staff surveys of initial order quantity and month. Since the salesperson is closest to the customer, they understand the retailer's requirements. They have a good idea of how many cases per store, quantity for retailer's DC (distribution center) and timing. They know the store plan-o-grams, resetting and buying behavior. These surveys work well because salespeople communicate their confidence about achieving the sale.
A key to success is to keep the process simple. We focus on initial orders because this is where salespeople add the most value. They work with the buyer to quantify retail plan-o-gram and inventory requirements. Avoid burdensome, overly complicated reports about too many issues. Eager marketing groups often request sellthrough and promotion participation. Forecast issues, not related to pipefill, are best handled through market research, consumption modeling and other communications. We simply ask for: the chances of getting into the account, the initial (pipefill) order if we get in, and the percent chance for each month. The survey is limited to the top 40 accounts targeted for the new product. Table...