Last Update: April 27, 2006




THE LATEST: RAMIFICATIONS OF ARKANSAS SETTLEMENT

After conducting further research into Google’s settlement of the Arkansas click-fraud lawsuit, we conclude that it is a much bigger win for Google than we initially thought. The settlement, if approved, will PRECLUDE ALL ADDITIONAL CLASS ACTIONS against the company based on PAST click-fraud claims. The California class action (see below), which was potentially far more serious than the Arkansas case, will disappear, as will all documents gathered from Google in discovery.

Attorneys for the plaintiffs in the California case called the Arkansas settlement “the worst class action settlement in history” [for the plaintiffs]. From Google's perspective, this suggests the settlement is the best settlement in history.

Because Yahoo! has not settled the Arkansas case, the California case against Yahoo! will continue, as will the one in Arkansas. Yahoo!, it seems, would be well-advised to settle the Arkansas case immediately.

Sources/Details:

We spoke with attorneys for the plaintiffs in the California class action. We also reviewed the Arkansas “Stipulation and Settlement Agreement” and a letter the California attorneys wrote to the judge in Arkansas after he denied their motion to intervene in the settlement. The letter lays out the settlement math, which will result in each plaintiff receiving not ‘pennies on the dollar’ but fractions of pennies on the dollar. Additionally, the plaintiffs will receive these fractions not in cash but in rebates against future advertising spending. Based on this, Google CEO Eric Schmidt appears to have understated the case when he called click fraud “not material” –at least as far as exposure from existing class actions is concerned. "Non-existent" might have been more accurate.

The attorneys in the California class action may continue to try to head off final approval of the Arkansas settlement, but are not optimistic about their chances of success. They will continue to press their case against Yahoo unless Yahoo also settles in Arkansas.

None of these events will preclude additional click-fraud claims against Google after the final settlement approval date.




PUBLISHED APRIL 24, 2006:

OUR VIEW

Since our last update, we have conducted significant additional research into click fraud (including several interviews described below). We continue to believe the click-fraud problem will eventually lead to higher costs for both search engines and advertisers. The more work we do, however, the less we worry that click fraud will cripple the industry.

We believe that advertisers will increasingly have to spend more money on auditing click-streams, thus modestly reducing ROIs. We believe the search engines will eventually be forced to spend more on auditing and customer service efforts, thus modestly reducing profit margins. We believe the problem is manageable, however. We do not believe that, say, half of all clicks are fake or that advertisers will suddenly boycott the medium en masse. Our sense is that, on average, click fraud is at the lower end of the commonly cited "10%-30% of all clicks" range.

The heart of the click-fraud debate is not about whether the problem exists (it does), but whether it can be managed. Click-fraud doomsayers believe that search engines can’t adequately police the problem because they don’t have data about post-click conversions (meaning that they can’t track what happens to a click referral after the click). As fraudsters get smarter about disguising fake clicks, this argument goes, the search engines will lose the ability to tell bad from good. The weakness in this argument is the fact that advertisers and third-party monitoring services can track click referrals through to conversion, and they can then confront the search engines with this data.

The search engines are currently reluctant to listen to advertiser complaints, and, sometimes, in a startlingly poor display of customer service, fall back on contracts saying they don’t have to. The engines will have to learn to listen and better address the problem, or they will frustrate customers and lose business. Reviewing and addressing customer click concerns will cost money, but it shouldn’t be a back-breaker. The cost for advertisers to audit click-streams, meanwhile, at least through one leading third-party analytic service, is less than 1%-2% of campaign spending. This cost is minor and could increase significantly without seriously affecting ROI.


OUR LATEST RESEARCH

In the past two weeks, we have conducted in-depth interviews about click fraud with several experts, including:

• A sophisticated advertiser who is convinced click fraud will destroy the pay-per-click business (and who has stopped buying Google keywords as a result). The interview was conducted on background; for details and notes, please click here.
• The CEO of a search-engine-marketing firm, Andrew Goodman, who agrees click fraud is a concern but does not think it is as serious as the media makes it out to be. Other topics included click fraud, search spending trends, local search, Yahoo, Google, and the (poor) outlook for Microsoft AdCenter. To read the complete interview with Andrew, please click here.
• The CEO of a sophisticated click-fraud auditing firm, Tom Cuthbert (see below), who believes the problem is overstated and can be managed. We will post the interview with Tom later in the week.
• A spyware and click-fraud expert, Ben Edelman, who has done painstaking research into the size and intractability of both problems. Other topics in the interview included spyware, Yahoo, Google, and porn ads. Ben’s recent treatise on the link between spyware and click-fraud at Yahoo is revealing and disturbing. To read the interview, please click here.

1. Economic Risk. A cottage industry has now sprung up to help advertisers monitor and analyze click-streams. Such services go beyond search-engine auditing efforts and give advertisers ammunition with which to request refunds. As more advertisers adopt such services, the cost of audits will impact ROIs.. Even if click-fraud is not rampant, therefore, the necessity of auditing click-streams will weigh on keyword prices. If click fraud proves a bigger problem than the search engines understand, increased advertiser scrutiny could result in demands for bigger refunds and improved screening efforts. Either or both would reduce search engine revenue and profitability.
2. Psychological Risk. Thus far, the click fraud problem has confounded efforts to quantify it. What’s more, click-stream analysis is subjective enough that advertisers and search engines may forever disagree about what is and isn’t a valid click. Third, researchers continue to discover ever-more-insidious strains of the plague, so just when the search engines have eradicated one, another breaks out. All this is leading to increasing FUD (fear, uncertainty, and doubt). One potential advertiser response is to stop spending money until the uncertainty is resolved. The lead plaintiff in a California click-fraud lawsuit, AIT, has done this. If the noise gets loud enough—and if Google, Yahoo!, et al, don’t do a better job of addressing advertiser concerns—this response could become typical. This, too, would hurt search engine revenue and profitability, regardless of how big a problem click fraud actually is.