How Wharton marketing students beat Wall Street analysts at their own game

This commentary by Peter Fader and Daniel McCarthy originally appeared on MarketWatch.

As notions of “customer centricity” gain legitimacy in commercial enterprises, it’s more important than ever to account for the number and financial value of a company’s customers. Corporate valuation can no longer be a purely “top-down” exercise without regard for the number and nature of the customers who create most of a company’s revenue.

Case in point: New customers acquired — the number of people who first bought from, or subscribed to, a company each quarter — has become a closely tracked investor metric in recent years.

‘Gross adds’

A growing number of Wall Street analysts track new customers acquired, trying to predict what it will be in the future. In a recently published journal article, we (along with London Business School professor Bruce Hardie) showed just how important the “gross adds” metric is to corporate valuation. Analysts who can predict how many customers will be acquired over time, and how long existing customers will remain with a company, will know the future size of a company’s customer base.

This greatly improves analysts’ ability to predict what future revenues (and thus profits) will be. We explain the entire process of going from metrics to stock price, which we call “customer-based corporate valuation,” in our journal article. Key headline figures such as gross adds already move markets to some extent, but we believe this trend is just getting started. Over the next five years, we anticipate that there will be a broad-based shift toward new ways of analyzing public companies that better leverage the underlying customer dynamics. In addition, we believe investors will begin to demand that all companies disclose customer metrics like those in a regular and timely manner.

Dish Network 

Consider Dish Network, the U.S. satellite-television provider. On May 1, Dish reported in its first-quarter earnings release that it had acquired 547,000 new customers. Wall Street was disappointed — this metric came in below analysts’ expectations, which contributed to net customers acquired and overall revenues trailing Wall Street’s estimates. Dish’s stock price fell that day and by the end of the week was down 4.4%, erasing $1.3 billion in shareholder value.

Wall Street analysts did not see it coming. Sell-side analysts’ forecast for gross adds as of April 12 (available through Thomson One) over-predicted by more than 13%.

Was anyone not caught by surprise? Yes, the students in our marketing course, “Applied Probability Models in Marketing.” Every year, we here at Wharton assign a project in which students apply the statistical models they have learned in the class to a real-world problem. This year, we tasked them to predict new customer acquisitions for Dish.

Accurate prediction

On April 5, almost a month before first-quarter results were released, every student submitted his best guess of gross adds for the first to fourth quarters of 2017. In light of Dish’s recent earnings release, the results are striking: a simple average of all 156 students’ first-quarter predictions is 550,000, a mere 0.5% from the actual additions. In other words, the marketing students trounced the Street. Those students have not spoken with company management, listened to company guidance, analyzed competing firms or studied historical financials. They simply came up with the best statistical models they could for gross adds, stress-tested those models on Dish’s historical data, then projected their models’ output for 2017.

The statistical building blocks of the students’ models are the same as those used in our aforementioned journal article on customer-based corporate valuation. They are so-called hazard models — models that predict the timing of events of interest. While hazard models are commonly used by actuaries to predict insurance claims, and by medical researchers to predict the onset of diseases, they are not used often by Wall Street analysts to predict the timing of new customer acquisitions. We believe that methods like those will begin to be adopted more broadly by the investment community as analysts start to think more carefully about the customer behavior “story” that underlies many of the numbers they are tracking or forecasting.

From bearish to bullish

In closing, let us return to Dish Network. What is in store for the second quarter of 2017? After a weak first quarter, analysts’ outlook is bleak. The consensus forecast for gross adds is 485,000. Our students, in contrast, are far more sanguine — the average of their predictions is 527,000, which is 9% above consensus. The students’ bearishness was well-founded in the first quarter, and despite the gloomy Wall Street outlook, we believe students’ bullishness in the second quarter will be equally well-founded.

Likewise, we are bullish about this idea of “customer-based corporate valuation.” No longer will marketers be relegated to focus on vague concepts such as “share of mind,” “brand equity” and “customer engagement.” Instead, they will become trusted advisers to the CFO, the Wall Street analyst and anyone who wants to understand, and effectively manage, the true value of a commercial enterprise.

Peter Fader is the Frances and Pei-Yuan Chia Professor of Marketing at the Wharton School of the University of Pennsylvania. Dan McCarthy is a doctoral candidate at Wharton, and this fall will be an assistant professor of marketing at Emory University’s Goizueta School of Business. The two are co-founders of Zodiac, a provider of customer lifetime value prediction and related insights.