Across almost every retail category, consumers are spending less and – perhaps more importantly – are in no mood to start spending more anytime soon. In a recent TNS Retail Forward ShopperScape study, half or more of the respondents indicated that they are buying fewer things, shopping less often and postponing purchases. If you are using predictive forecasting at the customer level, such models may no longer reflect today’s customer dynamics.
Start by asking yourself some key questions, such as: “How many “customers” do I no longer have?”, “How many customers have not made a purchase in the last 12 or six months?”, “How many of them should I stop considering a ‘customer?'”, and “Is it only the time between purchases that is lengthening, or I am permanently losing more customers?”
These questions have important significance for retailers trying to predict what future revenue to expect from their existing customer base, especially given current retrenchments in consumer spending. Lapsed or inactive customers will over-represent your forecast if included as current customers. Make sure that your projections are in fact based solely on customers still active with your brand.
Target the right customers. As part of forecasting customer spend, you also need to be able to predict the number and total value of purchases for each established customer over the next twelve months. With that analysis, you can then create a future purchase intensity score for each customer based on the forecast. But be careful: Once you have developed a scoring model to predict future spending potential, you need to resist the urge to just focus on the highest potential group. What will it cost to attract these customers to spend more with your store? Would it be more profitable to focus on customers with high (but not the highest) potential? Your sweet spot may not in fact be automatically the top percentiles.
Make the right offer. The final part of the analysis considers each customer’s receptivity to promotional offers, based on previous transaction activity, and assigns a responsiveness score. Together, these two scores can lead to a more targeted and timely set of marketing activities aimed at your active and established customer base. In short, if we can both estimate the timing of a customer’s purchase and know what offers will induce them to buy, we have a better chance of overcoming that customer’s caution caused by economic uncertainty.
Retailers can help turn around their sales and marketing performance by using forecasts of spending at the individual-customer level to drive a range of marketing programs across retail categories. Supermarkets, for example, can identify customer patterns for both immediate consumption and stock-up shopping and target Web site or e-mail promotions accordingly. Home improvement chains can find specific customer segments that will actually grow in revenue, in contrast to the overall softness predicted for the category in the immediate future. And apparel retailers, faced with predictions that nearly half of female shoppers plan to curtail their spending on clothing from a year ago, can focus on increasing their share of the “other half” of the market.
Ask yourself: What additional revenue opportunities could you create from targeting customers with high future purchase intensity scores? And, what improvement in response rates could you gain from using responsiveness scores to help determine targets for direct-response campaigns? Individual-level customer spending forecasts can answer these questions, and guide your marketing decisions to reach profitable customers in this challenging economy.