Archive for the 'Adcenter' Category

MFABlocker: New Service to Increase You Adsense Revenue

A friend of mine recently let me into an early beta of MFABlocker, a new service he is developing.

The Problem of Low Paying Clicks

Any webmaster or publisher who employs Adsense or any other form of PPC advertisements in order to monetize his sites has no doubt been frustrated from time to time with the miniscule revenues some clicks often generate.

The problem lies in the fact that much of the inventory advertising in Adsense is composed of MFA (Made For Ads) pages, which are often nothing more than pages full of ads. The people who run these sites will often bid very low on a large number of keywords in the hopes that people who click to their pages will click on higher paying ads. This is the concept of ad arbitrage.

While many consider this a legitimate form of business, there is no doubt that the practice warrants lower advertising revenues for the average webmaster.

MFABlocker - the Solution

Enter MFABlocker - a great tool to fight low paying clicks!

MFABlocker takes a list of your sites along with an Adsense publisher ID in order to determine which advertisers are showing ads on your sites. They then follow these ads, and analyze the resulting pages in order to determine whether they share any of the characteristics of MFA pages. If they do, they are added to a blacklist which you can then add into the Adsense competitive ad filter.

List of Blocked Sites in MFABlocker Approving Sites in MFABlocker

The Results?

So how well does it work? In the past week that I have been using this tool, I have seen my Adsense revenues jump by 25-30%. This is a welcome jump, although I admit that it may just be regular flux considering I have only been using it for a week. But there’s something, at any rate.

The Future

MFABlocker is still in early beta, but it is already showing some promise. The interface is quite rudimentary, and there is no help section yet. However, the service is developing rapidly. In the two weeks or so that I have been in the beta, I have been seeing major feature updates daily.

Keep an eye on this - it will be a great service!

Typo Squatter loses Thousands of Dollars Due to Missed Details

Update: the mystery is finally solved

Setting

This yesterday, I mistyped the URL as I was visiting Google this morning; I accidentally typed http://www.google.cm. This redirected me to a page on the domain of http://www.agoga.com, which actually looked like a somewhat convincing, spartan page, very similar in style to what you would often see if your browser. Except that it also contained a search bar, and a few unobtrusive links to subject like ‘Travel’, ‘Cars’ etc., the kind of subjects you would see on a typical parked domain page.

I thought that was kind of interesting, a way of monetizing typos that looked to me at least like it would be somewhat effective way of squatting a typo. At the time, though, it didn’t seem noteworthy enough to me to give it further thought.

A little later, I was trying to get to Paypal, and again I accidentally typed http://www.paypal.cm. Once again I was at the same page. I was intrigued, and began experimenting by checking a variety of other domains with the .cm extension. Many big names in the industry had the .cm TLD pointed to the same page I had viewed earlier.

That also, is not that notable. A squatter could easily have registered a whole variety of company names in that TLD - it’s done all the time, and is considered a valid tactic for making some money off of parked domains.

What made it notable finally is when I started entering random domains, and sequences of characters in the .cm TLD. such as http://sdfjhksd.cm and http://www.oiyt.cm. These also are pointing to a landing page on agoga.com, albeit a different landing page from the ones used on major domain mispellings.

Agoga.com has every unregistered .cm TLD pointed to their landing pages!

While there are a bunch of legitimately registered .cm sites which resolve elsewhere, any other .cm domain, whether nonsense characters or misspellings of ‘real’ domain names resolves to the same IP address which is a cluster at agoga.com. The only way this could be accomplished is to change the default site settings of the master DNS serving the .cm TLD. Agoga must have either hacked the .cm registrar in Cameroon, or paid the registrar off for this. Either way, I suspect something illegal has occurred here; I doubt this type of redirecting is approved by IANA.

Agoga Alexa Graph

Opportunity

How much type-in traffic would you think would be generated by people misspelling .com as .cm? Agoga.com has an Alexa Rank of 6,915 which indicates thousands or tens of thousands of visitors per day by some estimates. Keep in mind that this site has not been running for even three months yet; today’s Alexa rank was 2,913.

Since Alexa ranking is biased towards a technical crowd, I think it is safe to assume that the true numbers are fairly large. Now, it is easily attainable that a proper landing page optimized for Pay-Per-Click advertisements will result in a 30%-40% click-through-rate. Especially if one was to put some effort into ensuring the advertisements were targeted around the domain name or keywords at the similar .com page.

It is obvious that with this type of traffic, Agoga.com could be pulling in some huge advertising revenue - as much a $1000-$2000 per day. They should have it made in the shade, for all intents and purposes. But, they have screwed up royally.

How did they screw up?

Agoga will return you to one of two landing pages, depending on what type of domain you enter. One version, which they seem to use when squatting the domain of a large company or popular website, can be seen at . The other, which they seem to use for the domains of smaller websites and nonsense or misspelled domains can be seen at http://www.oeiurt.cm (note the random domain name…) or http://www.caydel.cm (a typo of this domain) or at the Agoga main page at http://www.agoga.com.

The first type of landing page is broken - The first type of landing page is relatively well done - it is minimal, and could easily get the user to click onto their main site. The problem lies in that no ads are served if the user enters certain search queries. While an advertising page is shown if the user enters a query such as ‘digital cameras’, ‘dvd’, ‘knitting’, other queries such as ‘infohatter’, ‘caydel’ or whatever return nothing. Sure, probably nobody is bidding on that term; wouldn’t it be a better plan to grab the first result from a Google query for that term, scrape it for keywords, and return ads based on that? Potentially millions of long-tail opportunities are being missed here, thrown away for no good reason.

The second type of landing page broken - The script that Agoga used to generate the second style of landing page is broken. Any search query or link click redirects you to the same page you just left, with a nice photo of a mountain range, or other scenery visible in place of the advertisements that should be shown. They are making nothing from this type of landing page; in fact, they are losing money due to bandwidth costs.

Opportunity Missed

I would be willing to bet that the majority of the traffic that Agoga.com receives will end up at the second landing page, the broken one. While they probably have their highest traffic domains such as http://www.google.cm pointing to their ‘working’ script, they are missing out on the whole long-tail of domain misspellings. Think about it this way - any mistake made by anyone anywhere when he misspells .com as .cm will send him to the broken script. This could be anyone typing in one of a billion domains.

Additionally, a fair number of people who misspell the the domains of large sites such as Google will make multiple mistakes - they may mispell google.com as google.cm, but how many are prone to make multiple mistakes such as gogle.cm or googel.cm and be sent to the broken page?

What Are You Trying to Tell Us Here?

The point of what I am trying to say should have become clear by this point, but I will write it out nice and neat anyways: an neglect of details can lose you a lot of money. I do not know if this second landing page has ever actually worked for Agoga. Perhaps it has, and only stopped working 15 minutes before I stumbled upon it the first time. Perhaps it has never worked. The fact of the matter is, the person or persons who own Agoga.com (Whois data indicates Nameview, Inc, BTW) are losing thousands of dollars per day. It is probably safe to assume that they don’t even realize this; if they did, they would fix it in realtively short order.
The people responsible for this had an amazing idea, which they ran with 90% of the way to the perfect money-making opportunity. But they have missed a few small details which are costing them perhaps thousands of dollars per day. If they were to fix these small problems, they could probably nearly double their income.

I appreciate your comments and feedback!

FOR THOSE COMING FROM REDDIT, READ THE EXPLANATION BELOW BEFORE YOU VISIT THE VIDEO:
If you look at the tags associated with Matt’s video here, you will see that someone has managed to get it associated with the Gay Porn tag. Is this a joke by an SEO or fellow Google Employee? Other attached tags include ‘Shoemoney was here’, ‘umpa umpa love’, ‘gilligan’ and more.

I think this is bloody hilarious. Obviously, someone has a great sense of humour, and couldn’t resist the opportunity to poke one at Matt. To whoever did this, great job!

Check out the

Rand Fishkin on Scraping and Search APIs

Last night, Rand Fishkin wrote an interesting article on the SEOmoz blog regarding the necessity of having proper search data and the inadequacy of the existing search APIs from the major search engines. He then explains the resulting need to scrape to get accurate data, and follows that up with the consequences of scraping, both for the scraper, and for the search engine being scraped.

I think that representatives from the big engines really should read this - it provides compelling reasons to update and ‘fix’ their broken/poorly implemented APIs. The data they can give is extremely valuable to companies, to the point that if necessary, many companies would be more than willing to pay for it.

At any rate, Rand provides a great listing of the important APIs, and great descriptions of their function. In addition, he makes some great suggestions about how such engines should be improved. Definately go read the article!

MSN AdCenter and Firefox

ShoeMoney has been playing around with MSN Adcenter and Firefox. MSN has been promising that they would have Adcenter supporting Firefox by August… It may happen, but why wouldn’t they write it to be compatible in the first place?

Never mind - we all know why. Same reason you can’t easily remove Internet Explorer from your Windows install…

Does this remind anyone else of the fiasco when people browsing with Netscape Navigator suddenly couldn’t enter ‘the  Zone’, Microsoft’s gaming site? I remember that. I also remember that you just had to modify the User Agent the browser sent to IE in order to gain access. Once you were in, it worked without a flaw…

The anti-competitive practices live on…