The above is an article of faith among a certain kind of Web publisher. And some of them are even paying for studies to prove that display ads–basically all the ads you see that aren’t part of search results–really do work on the Web.

Except when they don’t. That’s the unsettling conclusion that some research funded by Yahoo (YHOO) recently reached, the San Francisco Chronicle reports.

The study was produced by the Web giant’s Yahoo Labs, which has been getting new attention in the Carol Bartz regime and beefing up its staff of social scientists by “adding highly credentialed cognitive psychologists, economists and ethnographers from top universities around the world.”

One of the new hires, economics professor David Reiley, tried to track the benefits of a Yahoo ad campaign on behalf of a retail chain. He found that the ads did work, but only for people born before Woodstock:

The research, conducted in partnership with an undisclosed national retailer, sought to accurately measure the impact of Internet display advertising across online and offline sales, by tracking people who had registered with both Yahoo and the store. The research found an approximately 5 percent increase in spending among those who had seen the ads–with 93 percent of those sales occurring in stores.

The potentially worrisome thing, however, was that among those under 40, the percentage was nearly zero. That could reflect the unpopularity of the particular retailer among that demographic. Or it could underscore a growing immunity to display advertising among the Web-savvy younger generation.

Yikes. I asked Yahoo for its take on the study and the company sent me a (not surprisingly) sunnier summary of the research. Some of its highlights:

Major Findings:
By combining a controlled experiment with panel data on purchases, we find statistically and economically significant impacts of advertising on sales.

We estimate the total effect on revenues to be more than eleven times the retailer’s expenditure on advertising during the study.
93% of the effect was on offline (in store) sales.

Persistence: The effects of the campaigns were persistent over time, meaning that the sales impact could be tracked for a period of time after the campaign ended.

Demographics: there was no significant correlation or differences w/r/t location (by state) or gender.

But there was a significant difference w/r/t to age: customers over the age of 40 were significantly more responsive to the ads in terms of sales. The largest effect came from senior citizens (65+).

Clicks versus non-Clicks: Though clicks are a standard measure of performance in online-advertising, we find that online advertising has substantial effects on those who merely view but do not click the ads.

We find that 78% of the effect in sales comes from those who view ads but do not click them, while only 22% can be attributed to those who click.

Count me among the group disposed to think that brand ads on the Web do work, by the way. But then again, I have a vested interest in this being true since it’s what’s supposed to keep me clothed and fed. I’d hate to see scientific proof that it’s all a pipe dream.

For a contrary perspective, funded by people whose interests align with mine, check out this study funded by the Online Publishers Association.

[Image credit: pedrosimoes7]

About 1987 word got out that the Ingres database would soon have a sexy new function: It would be able to do distributed queries… Ellison told [Oracle ad man Rick] Bennett to prepare an advertisement announcing Oracle’s distributed capability. Then he assigned an engineer to whip up a distribtued feature so the company would actually have something to sell when the ad appeared. Ten days later Bennett’s advertisement hit the trade press: “Oracle Announced SQL*Star,” it said. “The First Distribtued Relational DBMS…”

“The fact of the matter was Oracle didn’t have anything,” said George Schussel, the trade show promoter who had followed Oracle from the beginning. “But that was the way they worked. Everything was marketing, everything was image. You simply announced the product and then figured out later how to deal with it from a technological point of view.”

– Excerpt from The Difference Between God and Larry Ellison

If you’re going to claim in an advertisement that Transaction Processing Council benchmarks show that a hybrid Sun-Oracle server runs faster than a competing product from IBM, it’s probably wise to make sure you have the TPC benchmarks to back up your claim.

Not if you’re Oracle (ORCL), though. On Aug. 27, the company ran an advertisement in The Wall Street Journal and the Economist claiming that “Sun + Oracle is Faster” compared to a TPC-benchmarked IBM (IBM) system. “Oracle and Sun together are hard to match,” Oracle said in the ad. “Just ask IBM. Its fastest server now runs an impressive 6 million TPC-C transactions, but on October 14 at Oracle OpenWorld, we’ll reveal the benchmark numbers that prove that even IBM DB2 running on IBM’s fastest hardware can’t match the speed and performance of Oracle Database on Sun systems.”

A boastful claim. Thing is, at the time it was made, the Sun (JAVA)-Oracle system hadn’t yet been audited by the TPC. In other words, it was based on an unsubstantiated benchmark. And that didn’t fly with the TPC, which fined Oracle $10,000 and ordered the software maker to pull the ad. “Oracle’s claim that it is faster than IBM using a TPC-C benchmark result it claimed would be announced on October 14, 2009 was not supported because Oracle did not have a TPC result at the time of publication,” the TPC explained in an official statement. “The TPC requires that claims based on TPC benchmarks must be demonstrable using publicly available data from official TPC benchmark results.”

About 1987 word got out that the Ingres database would soon have a sexy new function: It would be able to do distributed queries… Ellison told [Oracle ad man Rick] Bennett to prepare an advertisement announcing Oracle’s distributed capability. Then he assigned an engineer to whip up a distribtued feature so the company would actually have something to sell when the ad appeared. Ten days later Bennett’s advertisement hit the trade press: “Oracle Announced SQL*Star,” it said. “The First Distribtued Relational DBMS…”

“The fact of the matter was Oracle didn’t have anything,” said George Schussel, the trade show promoter who had followed Oracle from the beginning. “But that was the way they worked. Everything was marketing, everything was image. You simply announced the product and then figured out later how to deal with it from a technological point of view.”

– Excerpt from The Difference Between God and Larry Ellison

If you’re going to claim in an advertisement that Transaction Processing Council benchmarks show that a hybrid Sun-Oracle server runs faster than a competing product from IBM, it’s probably wise to make sure you have the TPC benchmarks to back up your claim.

Not if you’re Oracle (ORCL), though. On Aug. 27, the company ran an advertisement in The Wall Street Journal and the Economist claiming that “Sun + Oracle is Faster” compared to a TPC-benchmarked IBM (IBM) system. “Oracle and Sun together are hard to match,” Oracle said in the ad. “Just ask IBM. Its fastest server now runs an impressive 6 million TPC-C transactions, but on October 14 at Oracle OpenWorld, we’ll reveal the benchmark numbers that prove that even IBM DB2 running on IBM’s fastest hardware can’t match the speed and performance of Oracle Database on Sun systems.”

A boastful claim. Thing is, at the time it was made, the Sun (JAVA)-Oracle system hadn’t yet been audited by the TPC. In other words, it was based on an unsubstantiated benchmark. And that didn’t fly with the TPC, which fined Oracle $10,000 and ordered the software maker to pull the ad. “Oracle’s claim that it is faster than IBM using a TPC-C benchmark result it claimed would be announced on October 14, 2009 was not supported because Oracle did not have a TPC result at the time of publication,” the TPC explained in an official statement. “The TPC requires that claims based on TPC benchmarks must be demonstrable using publicly available data from official TPC benchmark results.”

The U.S. economy lost 524,000 jobs in December, closing out the worst year for job attrition since World War II, according to the BLS. Total job losses for 2008: 2.6 million, the largest decline since 2.750 million jobs were lost in 1945. A 16-year high. Congratulations, folks….

Suffice to say, that’s quite a bit more than some economists were expecting. And that’s an ugly, ugly number, 2.750 million jobs lost. With the national unemployment rate rising to 7.2 percent during December, the first quarter of 2009 is also looking pretty bleak. “The job situation is ugly and is going to get uglier,” Richard Yamarone, chief economist at Argus Research, told Reuters. “There’s no reason to expect hiring anytime in the next three to six months. We are not going to see any hiring until the government steps in and acts. Talk doesn’t work.”

Robert Barbera, chief economist at the Investment Technology Group, was even more pessimistic–if that’s possible. “I would suspect that starting this past October and lasting through April, we will have really big job losses,” he told The New York Times. “We are not yet near the numbers of those earlier recessions,” he added, referring to the downturns of the mid-’70s and early ’80s, “but five more months like what we have been having and we’ll be there.”

Despite copyright woes, it is getting kudos: The Economist has selected “Here Comes Another Bubble” by the the San Francisco-based Richter Scales as the best musical number on YouTube in 2007.

So, here is the video one last time: