Contextual ad shoot-out: Infolinks versus Kontera

I’ve had a very good relationship with Kontera and have been a premier advertiser through the use of its ad network on my Ask Dave site.
Many people in the online advertising space eschew the kind of contextual pop-up ads that Kontera run, however, because they say that it’s confusing to users: is a word in a different color, underlined, a hypertext link or a pop-up ad? Here’s how one looks:

kontera ad sample

My use on Ask Dave Taylor is perhaps even more confusing because I generally have it customized to be a single underline and to have the same color as a regular link (and my regular hypertext links are italicized, not underlined at all). Since the ad pops up on “mouseover”, I believe that it’s quite unusual for people to accidentally click on the ad. I will also say that I have served up millions of these ads since I started working with Kontera and never received a single reader complaint.
Other contextual ad networks have seen my busy site and asked me if I’d switch and generally I say no. The folks at persisted, however, and I decided it would be very interesting to test Kontera versus Infolinks on the site. They swore they could earn me significantly more money than Kontera was paying me and that sure got my attention.
I told each company they were going to be part of a test and that I would publish all the results, including actual earnings. Both companies agreed and were excited (and perhaps a little bit anxious) about the test.
I ran the A/B split test for the month of February, 2008. I used the following code to split ads 50/50 between the two networks:

<script type=”text/javascript”>
var konaPercent = Math.round(Math.random()*100);
if (adPercent <= 50) {
    // Kontera ad block
} else {
    // infolink ad block

The result? I’ve taken a 20 day slice to compare between the two, from Feb 5-25, and here first is the result from Infolinks:

infolinks report

The second result is from Kontera:
kontera report

Though you might initially think that the total amount earned is the most important, to me the key figure is the eCPM (effective cost per mil — where “mil” in this context is actually 1,000). With this figure you can see that Kontera offered up more valuable links: $4.39 versus InfoLink’s $4.23.
The two networks had exceptionally similar earnings: $1939 for Kontera and $1905 for Infolinks. Notice, though, that Infolinks had a higher click-thru rate (CTR) and more ad impressions to generate this slightly lower revenue when compared to Kontera.
The winner? Kontera, but only by a hair. We’re talking about a difference of about 1.8%, after all.
If you’re looking for an ad network to experiment with, Kontera might well be just what you seek. And if they don’t work out for you? Perhaps Infolinks would work instead.
Either way, my thanks to both companies for letting me run and publish the results of this test. I hope you find it interesting!

14 comments on “Contextual ad shoot-out: Infolinks versus Kontera

  1. Great analysis, Dave.
    Everywhere else I read, InfoLinks has earned more than double over Kontera. How can you explain that…? perhaps Kontera fixed their numbers?
    To be honest, what ECPM did Kontera earn you the same period of the previous month?

  2. Dave, great stuff. Thanks for sharing the insight and analysis. I am interested in how a service such as skimlinks, or skimbits would compare with Kontera or Infolink in your opinion?

  3. Good question, Vinay! In January my eCPM with Kontera was $1.69, quite a variant from these February results. Across the last six months, it was $2.37. I’ll ask Kontera to pipe in here with an explanation for the variation we see in these results.
    For the record, both Kontera and Infolinks told me that they’d push the best campaigns they could find to my site so that I’d have the best possible results, so I would be unsurprised to find that Infolinks has similarly skewed data, though I don’t know for sure.

  4. Thanks, Dave.
    Something is fishy. If Kontera gave you $1.69 ECPM and then it jumped to $4.39 only for the test, it means they gave you less then 40% revenue share by then. Was 40% your agreement with them?
    Perhaps we should all tell them that we test with InfoLinks and get double revenues?
    I will give it a try and update.

  5. Interesting !!
    I always thought Infolinks would be better than Kontera which I tried earlier.
    But now, I have to stick to the old one, Kontera

  6. That is one of the coolest post I’ve seen about monetizing your blog. I just clicked on a info link ad today and was in the process of filling out the application when I decided to do a little investigation first. I think I have heard of Konetra before but forgot about it. I signed up of for Chickta but the load time was insane. I am a new blogger with a low traffic rate so I like basing my decision off of facts and not opinions. Thank you for sticking to your guns so that others could make an informed decision

  7. Dave, I have a question about impressions and page views. Does net impressions in your comparison table mean page views or does it mean the popups? So effectively, do you get one popup opened per page view or do you get one popup opened for several page views?

  8. Great idea get the two company’s to agree to the test.
    Thank you for very accurate and well represented results.
    Interesting that they where so close. During the test do you notice any change it the location of the visitors that did click through?
    Thanks again for the info.

  9. I have website for over one year,made in Joomla,but Google AdSense program just dont want to work. I got acount after few days Google shut down adsense for no reason. I dont click on my site but…
    So me and my partner want ti switch to Kontera or Infolinks. Mr.Dave your test help me a lot. Regards

Leave a Reply

Your email address will not be published. Required fields are marked *