WHY IT'S SO HARD TO GET A READ ON AMAZON

WHY IT'S SO HARD TO GET A READ ON AMAZON

Amazon.com has a reputation for going to great lengths to help shoppers who visit its online store. But Wall Street analysts? Not so much.

Wal-Mart and eBay host analyst days to provide guidance on their financial outlooks. But not Amazon. Analysts almost never get to sit down with the company's top executives. And even as Amazon has gone from simple bookseller to behemoth with $19 billion in sales, it breaks out revenues in only three broad categories: Electronics and General Merchandise, Media, and Other. Tracking the Seattle giant has become like Internet-age Kremlinology. "We tend to find them generally unhelpful," says analyst Jeffrey Lindsay of Bernstein Research, "so we have to rely on our own ingenuity."

Last year, Lindsay wanted to find out if the retailer had a warehouse in Texas (to compute the effect of a state Internet tax). Amazon kept mum. So he recruited the Texas relative of a Bernstein employee to stake out a warehouse in Irving. Ultimately, the lookout confirmed the Amazon connection.

As the economy pitches and rolls, consumer spending is becoming a more important indicator of the country's well-being. So analysts are going to extremes to get a read on retailers. At traditional stores, some researchers lurk next to checkout lines to tally sales of spring dresses or flat-screen TVs. Others recruit students via craigslist to count the people going into malls and how many come out with bags.

But Web retailers, which account for a growing share of consumer purchases, present a special challenge: There are no physical stores to spy on. Wall Streeters tap independent researchers who run panels that track online consumers. But the results, like a lot of Web metrics, are open to debate.

TEA LEAVES

That has analysts concocting new leading indicators for Net retail. The latest experiment in tasseography (or tea-leaf reading) comes from Marianne Wolk, an analyst at Susquehanna Financial Group. She spent the past month Googling data online and poring over Securities & Exchange Commission filings to come up with a fresh approach: She's counting every square foot of warehouse space Amazon adds this year to see whether it can live up to Wall Street's financial expectations.

Wolk came up with the method after noticing that warehouse space is one of the numbers Amazon does regularly report in public filings. She went back to Amazon's 2002 SEC filings to tally up how much space the company has built, then compared that to the growth of Amazon sales and units shipped. She and her assistant also pored through newspaper stories to find out where the facilities were built and how big they are. "With a little digging, there's plenty of information," Wolk says. Based on Amazon's past expansion, she figures it's a bit shy of where it needs to be to meet consensus estimates for the second half of the year. She's watching for a ramp-up in capacity.

Amazon declined to comment on the report. Other analysts say the space-to-sales correlation may not hold up, if, say, Amazon gets much more efficient. Wolk will get an early indication of her accuracy on Apr. 23, when the company reports earnings. Meanwhile, Avi Cohen, head of research at Avian Securities, is dreaming up new tactics. Forget square feet, he suggests. Imagine instead that you could get college kids to lurk outside all of Amazon's warehouses to count trucks as they leave. Would it work? Analysts are willing to consider anything for an edge.