Monday, October 26, 2015

Things that a data-driven company ought to know

This is the first of a couple of Netflix stories I'd like to get to this week. Just to review, the price of the stock is very high relative to earnings and since Netflix isn't acquiring significant assets such as large content libraries and has no plans for (or apparent interest in) radically cutting costs, the only reason to buy the stock is a belief that the company's subscriber base is on the verge of explosive growth.

Under those circumstances, it's not surprising that the leadership of Netflix was scrambling to explain its failure to meet growth estimates recently; it was the explanation itself that caught analysts off-guard.

From Wired:
Netflix is blaming its lower-than-expected US subscriber growth to changes to Americans’ credit cards.

In a letter to shareholders, the company said its over-optimisitic estimates for its third-quarter results were “driven in part by the ongoing transition to chip-based credit and debit cards.” In other words, the company is claiming that the number of US subscribers to Netflix didn’t meet what the company had expected for the third quarter in part because, well, fewer people than expected paid up.

“I read this Netflix quote and I scratched my head and thought, ‘What?'” says Ken Oros, a senior associate at The Strawhecker Group, which focuses on the electronics payments industry.

In the past few months, credit card issuers have been transitioning from cards without chips to ones with them, known as EMV technology, to help curtail credit card fraud, which you may have noticed. Card users have been receiving these new cards in the mail as banks rushed to meet the October 1 deadline. Since then, new liability rules have taken effect that now hold merchants who don’t switch over to the new technology liable for credit card fraud.

To industry experts watching the country’s shift to EMV, Netflix’s statement is somewhat surprising. After all, while brick-and-mortar businesses have had to update their processing systems to account for the new cards, the way we pay at digital businesses remains pretty much unchanged.

“It sounds like a bunch of customers received new cards at once and their old cards on file were inactive,” says Forrester analyst Sucharita Mupuru-Kodali. “Most people may not even realize all the things they need to change for autobilling and they forget until the next time they use Netflix.”

Regular users, however, would likely soon realize if they weren’t able to sign into their accounts because they hadn’t paid their bill or if their cards were no longer active. “I can’t imagine it’s meaningful and, even if it is sizable for one quarter, I’m sure people will realize soon enough,” she says. “Anyone who churns out probably wasn’t using the service much in the first place.”


When pressed on the issue during the earnings call, Netflix chief financial officer David Wells clarified the statement from its letter about why US subscriber growth may not have met expectations. “We think it’s a contributor,” Wells said of the chip-card transition. “It’s likely multi-factored, there may be other things going on here, but certainly the transition to the chip cards is not helping and that has to be a factor.”

From the Wall Street Journal:
But those in the payments industry say their systems in place should prevent any such billing disruptions.

Henry Helgeson, who runs a company that processes transactions for small businesses, described the Netflix explanation as “curious” because other merchants haven’t complained about such a problem.

“I would be surprised if this was an issue in the industry right now and we’re only hearing about it from Netflix,” said Mr. Helgeson, chief executive of Boston-based Cayan, which was formerly called Merchant Warehouse.

Other payment industry experts also expressed doubt that the new cards were the root of Netflix’s disappointing subscriber numbers.

“If this was an issue, it would be affecting every subscription-based business and it isn’t,” said one card-industry executive at a large financial institution.

For subscriber based companies with automatic billing, the issue of churn based on credit card disruptions is a familiar problem, and, in those companies I worked for, it was very much a known quantity. Executives closely tracked these numbers on a weekly, and in some cases daily, basis. The thought of a C level executive from one of these corporations standing up and saying "we lost a bunch of customers last quarter. We think it might be because of credit cards." would be unthinkable. What's worse is that, not only do they not have solid data on this problem, their informal estimates appear to be wildly off-base.

For Netflix, this is particularly embarrassing. At the end of the day, the value of business data and statistics relies almost entirely on your ability to tie them to drivers of profitability such as acquisition, retention, pricing and cost. Netflix has aggressively and successfully pushed the narrative of being a heavily, even uniquely data-driven company, but, by the management's own account, they seem to be failing to collect the data required to take advantage of all of those online behavior metrics. Knowing how often a show is paused or viewed to completion is only useful if those metrics tell us something about retention and reactions to upcoming price increases.

As with most posts in the Netflix thread, the main significance here is not in the story itself, but in what its coverage says about larger trends. In the Twenty-first Century, journalists have become enamored with data, but their understanding of statistics has not significantly improved and the results are, in practical terms, worse since common sense is increasingly pushed aside to make way for misinterpreted numbers and badly understood technical terms.

In addition to this lack of analytic understanding, the Netflix story also plays on other longstanding weaknesses of the press: an appetite for simplistic narratives (especially ddulite narratives); an infatuation with visionary CEOs; a tendency to defer to authority. Fortunately, with Netflix, the stakes are fairly low and gullible journalists can't cause that much damage. Unfortunately, these same weaknesses apply to journalists covering segments of the economy like financial services, and there the stakes are not low at all.

In case this sounds familiar.

No comments:

Post a Comment