Call it the new opacity.
Despite the best efforts of research firms to pump out data that seem to forecast how marketers are doing, complexity is growing faster than the ability of research firms to cut through it.
|DATA: Package-goods market share, and ComScore data vs. Google sales|
The most notable example is the gyration of Google's stock in recent months (down more than 40%, then back up more than 30%) after U.S. paid-click data from ComScore seemed to indicate a slowdown that didn't materialize in the search giant's reported quarterly earnings.
Sanford C. Bernstein & Co. analyst Michael Nathanson last month detailed how many streams of data he tracks regarding the performance of media and entertainment players -- and how little the data seem to reflect marketplace performance. "The unintentional game of bait and switch that just occurred with Google is unfortunately becoming an all-too-familiar occurrence to media investors," he wrote in a note last month.
Reliable data needed
The increasing importance of global markets to almost all marketers is also creating more nooks and crannies where the bright light of number crunching never shines, with key growth drivers such as China, India and Latin America providing little in the way of comprehensive, reliable retail data.
For most package-goods players, such markets are growing three to five times faster than developed markets where data are more readily available, and they make up anywhere from a quarter to nearly half of global sales.
Combine that with a trend toward major players giving no quarterly earnings guidance (Google) or far less guidance (Procter & Gamble Co., which this year stopped providing mid-quarter updates), and you have the makings of increased market volatility -- maybe forever.
"It adds a ton of volatility to the stocks," said Bernstein consumer-products analyst Ali Dibadj. "It also allows the companies to be a little less transparent with what they do to make their numbers."
Market shares in many developing markets are largely educated guesswork. This has fueled a sort of Lake Wobegon effect. "What you're seeing now is all these companies saying they're gaining share [at the same time]," Mr. Dibadj said.
Not on same page
A better indicator may be comparing organic sales or volume growth among the major global players. But that, too, depends on marketers all applying the same yardstick for a metric that isn't regulated by accounting-standards bodies.
Colgate-Palmolive Co., for example, adds the impact of "mix" (consumers trading up to higher-value products) to its volume, Mr. Dibadj said. And laundry manufacturers' concentrating their detergents has rendered volume numbers less meaningful too.
On a smaller scale than Google, the stocks of consumer-products marketers have been buffeted by data disconnects in recent months.
P&G's stock had a brief relief rally late last month after its quarterly results outperformed downbeat "whisper numbers" fueled by what scanner data in the U.S. and Europe indicated would be a weaker quarter.
Estée Lauder last week missed its fiscal-third-quarter earnings target, yet its stock soared 7%, in part in relief -- and covering of short positions -- following whispers of an even worse quarter based on department-store data from NPD Group and weak beauty results for P&G and L'Oréal.
Reckitt Benckiser surprised analysts and the market last month when its organic sales growth of 11% beat consensus analyst projections of 7.6%, which were based heavily on U.S. and Western European scanner data. Likewise, Kimberly-Clark Corp. has substantially outperformed predictions based on scanner data for the past three quarters -- leading to never-confirmed speculation of a big, new private-label account.
U.S. scanner data for such players hasn't covered Wal-Mart Stores since the retail giant stopped giving numbers to Nielsen Co., Information Resources Inc. and NPD in 2001. Still, analysts relied on the notion that changes in market share in the tracked channels at least reflected broader market trends.
But it's no longer clear that's true, particularly as a slowing economy has driven more consumers to shop at big-box stores with more-limited assortments. Costco, for example, consistently has been the fastest-growing major retailer in the country, but Costco sales do not show up in ACNielsen or reported IRI data (IRI sells scanner data from Costco, but not to analysts). Not only that, but success at Costco could hurt shares in the tracked channels to the extent that brand-loyal consumers shift their purchases there.
In the media and entertainment world, it's no better. ComScore, NPD, IRI, Nielsen Media, Smith Travel and Arbitron "have generated increased revenues by selling their products to data-point-hungry analysts," Mr. Nathanson wrote.
Unfortunately, all those data streams have proved to have little bearing on financial results or stock prices and sometimes become counter-indicators, such as in the case of Orlando hotel-bookings data (which exclude Disney properties), he said. Viacom and CBS, for example, both have shown remarkable ability to overcome declines in ratings either through raising prices or selling more ads, he said.
Mr. Nathanson said he still believes the data can spot macro-trends, but it can't necessarily predict quarterly results.
For its part, ComScore went to great lengths to show how its paid-click data correlate rather closely with Google's U.S. search revenue. The disconnect with the market, said Chairman Gian Fulgoni, came because some analysts put too much stock in the data.
Thanks to Google's quality-scoring system, even U.S. revenue doesn't align exactly with U.S. paid clicks, Mr. Fulgoni said. And, as with other U.S.-based businesses, faster-growing international markets are driving an ever-larger portion of Google's revenue.
"Market research is taking on some really complicated challenges," Mr. Fulgoni said. "It's really complicated to try to track all the paid clicks everywhere they're happening." ComScore, he said, is the only provider that even tracks U.S. paid clicks, and it began doing so only recently.
IRI and ACNielsen, too, have made household-panel data tracking Wal-Mart sales available to analysts, though not all buy it or share it with clients. And IRI is attempting to cut through the increasing need to merge market data from multiple sources with a new "liquid computing" platform, though analysts don't necessarily have all the data sources (such as Costco scanner), much less the platform.
Beyond that, investors may have to permanently come to grips with the fact that the numbers they can see don't mean much. "The market is starting to catch up with the fact that this stuff is more and more irrelevant," said Bernstein's Mr. Dibadj.
The problem is, it isn't always irrelevant, said William Leach, a buy-side analyst with Neuberger Berman. Energizer had a bad quarter recently, "which was sort of tipped off by the scanner data, which showed the battery business declining," he said. "Campbell Soup had weak sales [indicated by scanner data]. If [the data] was always bad, you could just ignore it."