New Orleans–Following a Saturday night on Bourbon Street, a Sunday morning session on merchandise forecasting might have been too much to process. But the panelists at the session “Forecasting the Future” at the DMA 87th Annual Conference and Exhibition here managed to make the topic accessible and even enjoyable.
Among the suggestions offered:
* When creating a response curve for a catalog mailing, be sure to include a Week 0, said Gina Valentino, vice president/general manager of Shawnee Mission, KS-based consultancy J. Schmid & Associates. “Magically, orders come in even before the catalog has mailed,” she said, and your curve should reflect that.
* A standard six-week order curve will typically peak at Week 3, and roughly 85% of orders will be received by the end of Week 4. But sale books and holiday remails will typically spike sharply during Week 1 and then almost as sharply decline during Week 2.
* Prospects generally react more slowly to catalog mailings than do customers, who already trust the brand, Valentino said. So if a mailing is going to a higher-than-usual percentage of prospects, you may need to shift your order curve and its peak accordingly.
* While historic trending and purchasing patterns per mailing are of course critical when forecasting demand for future mailings, numerous considerations can create variances in actual vs. expected demand, said Margo Wycoff, vice president of planning and invenotry control for Boca Raton, FL-based women’s apparel marketer Boston Proper. For instance, if an item was featured on the cover of a previous catalog but is within the well of the book this time, demand for that item will be less than you would expect based solely on the numbers from the previous mailing. In addition to placement in the catalog, seasonality, creative treatment of the item, price changes, and where the product is in it life cycle also could influence demand projections.
* Natural disasters (the Florida hurricanes, for instance) and major national or world events (the Olympics, a terrorist attack) that suppress demand can’t always be predicted, of course. But you should take note of any occurrences in your forecasting records so that when projecting for next year, you don’t inadvertently underestimate demand.
* The average price point offered vs. the average price point sold is a key criterion to track, said George Mollo, president of merchandising and operations consultancy GJM Associates. The difference between the average price point offered and the average price point sold should not exceed 20% in either direction. If the price point sold is significantly lower than the price point offered, you may need to offer more lower-priced goods. Conversely if the price point sold is appreciably higher, you may be able to introduce more higher-priced items into your merch mix.