• by Ray Schultz , Columnist, December 11, 2019

First impressions matter — especially in retail. And they can affect sales more than other variables, judging by a paper highlighted this week in the Harvard Business Review.

We’re not talking about impressions absorbed by the customer coming in, but about those gleaned by the company when the first purchase is made.

Based on those insights, a firm can “separate high-value customers (from those who are unlikely to purchase again) as well as identify those who are more sensitive to marketing interventions such as email campaigns,” write Nicolas Padilla, a doctoral candidate in marketing at the Columbia Business School, and Eva Ascarza, the Jakurski Family Associate Professor of Business administration at the Harvard Business School.

The authors have created a model that they claim can help brands improve their marketing based on what they learn from the first transaction.

Their algorithm “creates a first impression of each customer much the same way two people ascertain each other’s character on a first date,” HBR wryly notes.

What’s wrong with traditional marketing metrics that accumulate over time? 

Well, for one thing, recency does not distinguish first purchasers because “all were acquired at the same time,” the authors note.

And demographics such as age and gender are not always available at first touch. But brands can observe other behaviors — for example, was the purchase “online or offline, did she use a tablet or computer, did she buy a new product or an old best-seller, did she buy on discount or at full price”?

Case in point: An online buyer “is more likely to have been acquired online (compared with customers whose trait is low), but also might be more receptive to promotional emails sent from the firm, once the customer has been acquired,” the writers continue.

They add: “We assume these traits to be unobserved to the firm, and stable over time."

Padilla and Ascarza studied the purchasing patterns of 13,473 customers of an international retailer over four years.

The authors conclude that the “firm’s email campaign sent to newly acquired customers would have increased its ROI by 39.7% had the firm targeted those new customers based on their first impressions obtained from our model.”

Brands seeking to exploit these findings could ask whether shoppers attracted by promotional emails for cheap or heavily discounted products? They will also be influenced by discounts when they repurchase — if they do at all.

Of course, consumer behavior is also influenced by “market conditions and seasonalities, promotional emails targeting specific customers and random factors unobserved to the firm at the moment of purchase,” the authors note.

We assume that many savvy marketers know these truths. How well they apply them is another matter.

Reading the paper might be a good beginning.