Skip to main content
ABC News
Black Friday Sales Numbers Are Useless And Wrong

Sometime on Thursday, after the football game and the second slice of pie, you might turn on the television and see a cheerful local news reporter standing outside a mall and reporting on the long lines of deal-seeking shoppers. Or maybe it will be a Friday morning headline proclaiming, “Black Friday crowds signal strong holiday season.” Or perhaps the news will be grimmer, and the headline will read, “Sparse crowds worry retailers.”

You should ignore these stories.

On Sunday and Monday, the stories will become more definitive, with actual numbers. They’ll say that Black Friday sales rose 2.5 percent from a year earlier, or fell 5 percent. They’ll quote experts on what the weekend sales meant for retailers, for the economy, for the election.

You should ignore these stories, too.

It’s understandable that journalists, economists and other observers would want an early peek into Americans’ holiday shopping plans. Consumer spending makes up roughly 70 percent of the U.S. economy, and the holiday season accounts for a disproportionate share of that spending. Nearly a fifth of retail sales come in November and December,1 and for many specific sectors, the share is even greater. Consumer spending may be even more important to the health of the economy than usual this year, when other sectors of the economy such as manufacturing have shown signs of slowing down.

Unfortunately, preliminary Black Friday reports contain almost no useful information about the state of the economy. The first round of stories — the ones on Thursday evening or Friday — are driven almost entirely by anecdote: How many people showed up at the mall where a TV crew happened to set up its cameras? How long do the lines seem to be compared to last year, according to a mall manager with a vested interest in making sales look good? How much was spent by the handful of families willing to talk to journalists on their way back to the car? These stories help fill airtime on what is typically a slow news weekend, but they aren’t good for much else.

The second round of stories, on Sunday evening or Monday morning, at least attempts to move beyond anecdote into real data. But as blogger and investor Barry Ritholtz has argued for years, early Black Friday sales figures are at best unreliable and at worst completely useless for predicting overall holiday sales.

The most widely cited numbers come from the National Retail Federation, a trade group that commissions an annual survey of Thanksgiving-weekend shoppers. Last year, the NRF said sales fell 11 percent from the previous year. When the Commerce Department released its final numbers for the holiday season, however, it found that retail sales rose about 4 percent in November and December from a year earlier. It’s hardly surprising that the NRF’s numbers, released on the Sunday after Thanksgiving, don’t accurately predict the holiday season. First of all, it’s difficult, maybe even impossible, to produce reliable figures that quickly. It takes the government weeks to produce even preliminary data on sales, jobs or other major economic indicators, and those numbers are revised as more complete information becomes available. To get its numbers out in time, the NRF conducts its survey before the Thanksgiving weekend ends and asks respondents how much they expect to spend over the full weekend.

Other sources came up with estimates that were wildly different from NRF’s last year. ShopperTrak, an analytics company that collects data on in-store sales, said sales fell 0.5 percent. MasterCard, which collects data on spending using its payments network, said retail sales had risen 2.3 percent in November (3.6 percent excluding gasoline sales). Those sources aren’t all measuring exactly the same thing, but it’s a bad sign that they’re pointing in such totally different directions.

NRF spokeswoman Kathy Grannis Allen said the organization stands by the accuracy of its estimates. But she said the headline-grabbing sales figures aren’t necessarily the most important part of the survey, which she said has successfully picked up trends in the past such as the shift to Thanksgiving Day shopping and the rise of online retail.

The second problem with the numbers is even more fundamental. Even if we could know exactly how much consumers spent over Thanksgiving weekend, we still wouldn’t know what that meant for holiday shopping overall. Holiday spending patterns have changed significantly in recent years. MasterCard last year said Dec. 23 was a bigger shopping day than Black Friday. Consumers are doing more of their shopping online, and retailers are competing to offer deals earlier. Many stores now start offering their “Black Friday” deals online the week before Thanksgiving.

The changing patterns make sales figures hard to interpret. If Black Friday sales are down, is it because consumers are feeling less confident or because they’re waiting to do their shopping later in the season? If online sales are up, is that a sign that overall spending is rising, or is it just the result of consumers doing more of their shopping on the Web? Shopping patterns are too unstable to know for sure. The NRF itself cautions against reading too much into Black Friday sales numbers.

“The weekend is tremendously important for retailers, but in terms of it being an indicator for how the season is going to perform, it’s not the best bellwether,” the NRF’s Allen said. She said a weak economy might actually lead to more Black Friday shopping, as deep discounts appeal to cash-strapped shoppers.

Perhaps in time, consumers’ behavior will settle into a fairly reliable pattern that’s easier to predict. For now, though, the only way to know how the holiday season will go is to wait and see.


  1. Excluding money spent at gas stations, building and garden suppliers, and auto dealers, 19.9 percent of retail sales last year took place in November and December.

Ben Casselman was a senior editor and the chief economics writer for FiveThirtyEight.