Quality Score Update – CPC to PPC Arbitrage Parade

In April, I wrote about a site that had jumped from zero traffic to an Alexa reported almost 15 million page views per day. The site peaked towards the end of June at more than 20 million page views and achieved a rank that would make many traditional publishers envious. In the “traditional” realm, sites receive traffic from a variety of sources including organic search traffic, referrals from other sites, repeat visitors, and paid ads. What makes sites like the one profiled in April unique is that they receive 99% of their traffic from paid ads.

Receiving one’s traffic almost entirely from paid ads doesn’t necessarily imply anything unusual or negative. Many lead generation sites that run on search receive a significant percentage of their traffic from paid traffic. What separates those sites from the group being illustrated here is that one is an ad; the other is ad supported. A form for a mortgage quote is an ad. Yahoo is an ad supported site. CPC to PPC arbitragers, one of which being 8bestsites, not only receive their traffic from paid sources, in this case exclusively search, but they also take that traffic to a page of ads, not a single ad or content with ads. This is why in my original story about arguably the largest of such sites, 8bestsites, I described the process as robbing Peter to pay Paul.

8bestsites and those like it have a straightforward model. They place ads on Google and then send traffic to other advertisers on Google via Google’s affiliate program, AdSense. This is to say that click arbitragers or click moppers as I like to call them simply redirect a user from one paid ad to another. The first paid ad is their own that they place on the engine. The second paid ad is a click on a similar paid ad where they are the affiliate. They make their money by buying lesser valued terms and sending users to a page with higher paying terms.

An example would be that they place an ad on Google for $.10. A user clicks on that ad and goes to 8bestsites where they see more pay per click ads. Here though, the average click is $.60. If 8bestsites see one out of every six people clicking, they have broken even. If they see anything better than that, e.g. one in four clicking they have made $.15 when they only paid $.10. Do this across a wide range of keywords, and they can generate significant revenue.

As is the case with many ad revenue streams that generate high returns with little user value ad (think untargeted emails, adware), the fun will ultimately come to an end – either through legislation or proactive business practices. With respect to the click arbitragers, Google has begun to put their foot down. In their Inside AdWords update email sent July 10, 2006, the company references an upcoming change to a particular algorithm used when determining minimum bid prices for their advertisers. (Unfortunately or perhaps typically, they provide no date for the algorithm update.)

In the email, Google says that “this change to the algorithm will affect a small number of advertisers” and that Google would “rather show one less ad than to show an ad which leads to a poor user experience — since long-term user trust in AdWords is of overarching importance.” The email does not provide insight into which “small number of advertisers” will be affected, but ClickZ shed more light by suggesting that the algorithm change would take a tougher stance on “made for AdSense” sites, e.g. 8bestsites. In shrewd fashion, Google won’t ban them; they instead have borrowed tactics used by auto manufacturers for stocking parts to older cars. They will raise the rates to price the owners of such sites out of the market. It is forced evolution that will not impact even if the change results in their extinction. The sum total of all CPC to PPC arbitrage, most likely, does not exceed $10 million monthly. I suspect that they won’t go away; they will simply trim in numbers.

Presented below are examples of click arbitrage. (Note: picture quality for some is not that great.) As I see it there are three categories of click moppers – 1.) No Content, 2.) Domainers, and 3.) Some Content. We begin with those presumably most targeted by the new algorithm change.

Fig. 1.1 – the almost infamous 8bestsites who bids across tens if not hundreds of thousands of words.

Fig. 1.2 – Perhaps the same group as 8bestsites or just their protégé, Top4Sites does not have the same coverage. Interestingly, performing a query by entering a term into their search box returns high CPC results unrelated to the user’s request. Only at the very bottom, well below the fold do the requested results show.

Fig. 1.3 – Another quite literal site, listm.net (List them?) also displays a healthy does of AdSense ads as its sole content. This site appears smaller than 8best and Top4.

Fig. 2.1 – Here we see an example of the second category, Domainers. These are people who specialize in the ownership of domain names; in this case it’s a marquee name. Domainers excel in the acquisition of Internet real estate but look to others for helping yield returns on that name while they wait to sell it or develop. When set up properly, traffic can be highly targeted and convert well. The ads that appear serve as placeholder content that often direct users who have typed in the site to reach a site that might satisfy the intent that drove them to the domain in the first place.

Fig. 2.2 – Here we see another example of a domain name holder driving traffic to their AdSense supported parked home page. This site has slightly enhanced functionality over the last example. Most interestingly though, the owner of this site is a public company, one that owns hundreds of thousands of domains. You have to wonder how much of their revenue comes from arbitrage, whether the investors know, and what impact they might see as a result of this change.

Fig. 3.1 – Here we have an example of 8bestsites +1, that is a site which does its best to look closer to a fully functioning site even though it falls squarely in the made for AdSense category. Users are focused and directed to the ads at top which dominate the leaderboard real estate and the first 30% or more of the page.

Fig. 3.2 – “Where’s the info” indeed. This site contains almost nothing that will ultimately help the user find what they came here for originally. As is the case with the other sites, there is nothing explicitly wrong. They are simply taking advantage of pricing discrepancies and working the difference in their favor…for now at least.

Fig 3.3 – Last and perhaps anything but the least we see a much more technically savvy site than those that preceded it. Containing more info than where’s the info (sounds like a horse name), http://www.free–ringtones–online.com is built on a simply, but more sophisticated data base driven content schema.

Links on the left change the content and as a result the ads. The site can test to see which of its content pages works best with what keywords, with a few of its content pieces bordering on helpful. A lack of updating and being too templatized should ultimately hurt sites like this.

If you see any others, do share.