Hear No Evil, Say No Evil, But Believe People are Evil
|
Written By Reprise Media | November 15, 2005 | Share This
|
|
It seems as though everyone’s favorite topic these days is “The Long Tail.” And while I’m a little burned out on all the discussion about it myself, the case could be made that I’m really no exception.
After all, in my most recent SearchViews entry, I simultaneously applauded and questioned Yahoo Search Marketing’s decision to access the long tail of advertisers by eliminating the $20 monthly minimum associated with its paid search programs.
On the one hand, YSM will be increasing the number of advertisers using its service. But the fact that its minimum CPC still stands at $.10 - and Yahoo seems unwilling to drop it - means that YSM isn’t using the same logic to access the long tail of keywords. Increasing the percentage of queries that it monetizes would seem to be the easiest way for YSM to drive incremental revenue from its search service.
Google, on the other hand, has become the poster child for long tail companies (some of its employees even went so far as to dress up as The Long Tail for Halloween). The company has built one of the world’s most profitable businesses (in record time, I might add) by providing advertisers too small to warrant a phone call from a sales rep with an automated means of reaching a massive web audience.
And its AdSense product has provided web publishers who could never hire an internal sales force to sell their own inventory with a revenue stream that can be plugged into a site in a matter of minutes. In contrast to Yahoo, the minimum CPC across AdWords has been $.05 for some time - thereby encouraging advertisers to create massive keyword lists that are inclusive of “tail terms.”
As Chris Anderson has pointed out, nearly half of Google’s revenues are now nearly split between ads served on its search results pages and ads served on content pages. And approximately half of the revenue it generates from advertisers is made up by those who could be classified as “long tail.” Google touted these accomplishments when it opened its February 10 “Analyst Day” presentation with a slide that read, “Serving the Long Tail.”
But while Google’s focus on the long tail has created benefits for its customers (i.e. advertisers, publishers, and consumers) and the company itself in the past, Google has made a couple of recent decisions which lead me to believe that its much more focused on profitability at this stage. And that profitability is ultimately going to come at the expense of its advertisers.
No API for You!
Any search marketer managing large-scale campaigns will tell you that one of the keys to its success is API access with each of the major search engines. If an advertiser can change her bids, keywords, match types, or any other campaign variable more frequently than her competitor can, she’s gained a significant advantage.
And generally speaking, the efficiencies that these more “sophisticated” advertisers gain through ongoing access to an API lead to substantial increases in spending within the search marketplace. So, the advertiser benefits from a higher volume of customer acquisitions and the engine benefits from the associated increases in spend.
Recognizing that this was the case, Google decided against charging its advertisers for access to the AdWords API when it was released earlier this year. Instead, it implemented a “token” system that was based on an advertiser’s monthly spend. Simply put, the more you spent, the more access you had. And most advertisers and SEM’s had enough tokens to meet their needs.
For reasons only Google can explain, the company recently decided to start charging for access to its API or usage above and beyond the allotted monthly quota. In most cases, this will only affect large advertisers and SEMS that have invested resources and time to build tools and technologies to leverage the efficiencies the API provides.
However, for those marketers who recognize the incremental opportunities achieved through the long tail, this API fee essentially translates into a highway toll on the management of increased keyword volumes in the long tail.
While it’s understandable that the tail does represent an increased load on technical infrastructures, said marketers cannot exceed the savings they put into the bank from less active account management and sales support or the email costs of passing large 65,000 keyword media plan files.
The Bottom Line
Charging advertisers - even once every 24 hours - for access to a report that contains thousands of zeros in both the click and impression columns is going to be a disincentive for those who were previously creating and managing large-scale campaigns. Tail terms that were previously profitable at $.10 or less now carry a cost even when they haven’t been searched for!
This could very well lead to a decrease in the percentage of queries that Google monetizes. More notably, API charges will prevent advertisers who were overspending on their core search terms by using tail terms to bring their overall CPA’s within reason from spending as aggressively as they have in the past. Not only does that mean less volume for marketers, it means decreased revenues for Google.
Enter Quality Score?
Most AdWords customers haven’t been using the service long enough to know all of the pricing changes that have been made over the years. That history goes something like this: the two “featured” spots at the top of the page were previously sold on a CPM basis, while the ads on the right side of the page were subject to the open CPC auction. But while the minimum CPC for those ads was $.05, Google had assigned arbitrary “floors” for what it considered to be “premium” terms.
So, the minimum CPC for a keyword like “flowers” might be as high as $.40/click. Over time, Google came to its senses, recognizing that the market would always find (and often exceed) fair value on its own. All of its placements, including the “featured” spots were placed in the same auction with $.05 minimum bids for every keyword.
Ad rank was ultimately determined by a blended metric of maximum CPC x click-through rate. Google was simultaneously rewarding advertisers with more relevant ads with higher placement (while requiring lower CPC’s) and guaranteeing themselves that they were giving priority to the ads that would deliver the highest effective CPM.
And the model obviously worked. But as opaque as it was, perhaps it had too much transparency for even Google. They recently amended the system yet again and have moved towards a “Quality Score” system. For those not familiar with Quality score, it allegedly seeks to reward ad relevancy to an even greater extent than AdWords did in the past. Rather than simply looking at click-through rate as a measure of relevancy, Google now takes a variety of other factors including ad copy and visible URL into consideration (of course, there’s no transparency with Quality Score, so it’s impossible to say what else they may be looking at or exactly how each of these variables is weighted).
More importantly, Google is no longer evaluating the merits of an ad on its own terms and based on its own performance history. Rather, each purchased keyword is assigned a score the moment it’s implemented. This score is based on the aggregate performance of all advertisers who have purchased that same keyword over time. So, if an advertiser bids on a keyword that has historically performed at an average CTR of .1%, the system will assume that the new ad will perform the same way.
What are the implications of this? For one, new advertisers, who have not established a Quality Score of their own will have to launch with increased bids to outrank their more entrenched competitors. But what’s more significant is that Google takes the same approach to Quality Score even when there’s only one advertiser bidding on a given keyword.
For example, let’s say my firm decided to buy the term “Reprise Media” (which we actually do) and there were no other advertisers bidding on it (which there aren’t), Google would still assign us a Quality Score based on the performance of every other company who had purchased “Reprise Media” in the past. And because it’s unlikely that any of those ads performed particularly well, we’d need to pay more than the alleged minimum bid (which was recently changed to $.01 with the launch of Quality Score) for our ad to be served. In fact, it’s possible that the CPC we would have to pay to buy our own brand name would be closer to $.50/click. As the ad began proving its relevance with a high CTR, etc. the CPC we’d be required to pay would come down accordingly.
Needless to say, this doesn’t bode well for those advertisers who have leveraged the long tail in the past. Generally speaking, these are keywords that get relatively few searches on a monthly basis, but are extremely profitable because they are typically bought for pennies a click. Google is now asking advertisers to potentially lose money on ads as they prove their relevance.
Once advertisers get their heads around the concept of Quality Score, fewer and fewer of them are going to be prepared to do that - particularly when there’s no one else bidding on a given term. As a result, the long tail that Google’s been making a point of serving over its brief history will be getting shorter and shorter.
Google may not be evil, but judging by its Quality Score, the company certainly seems to think that people (well, at least search marketers) are inherently corrupt.
Peter Hershberg is Managing Partner of Reprise Media.
Topics: Search: Innovations |