Cash is Good!

I have taken a part-time job, and it's definitely affecting my blogging time. I'll continue to post here as I am able.

Friday, November 27, 2009

Have I Caught Entrecard Cheating?

I wrote at the beginning of November about the results of some experimental clicking to see if Entrecard was delivering only the 15% of ads for their system which they claimed. Using the methods that I chose at that time to load the blog pages, the ad rate did work out to 15%.

But some questions were raised about other methods of serving blogs, so I decided to so some further research today. It was quite enlightening.

First I opened 100 blogs in the Entrecard system by using the EC toolbar. My selections for the type of blogs to open were from several categories that interest me, and the list was further filtered by "Most Popular." I often click this way, knowing that those bloggers are likely to reciprocate and visit my blog. For 100 blogs, there were 16 EC ads. This is certainly within the margin of error, although it is interesting to note that all 15 were served by the time I had opened 80 pages. In the 100 pages opened, 2 of them had no widget, which I thought was pretty interesting for a blog that had made it into the top 20 in a category (I chose 20 from each of 5 categories). I did reload each of those 2, to be sure that the widget really wasn't there.

But when I switched strategies, I really had an eye-opening result. I did two more sets of 100 blogs each, served by the EC widget. But for each of these sets I chose All and Random for the parameters of the search. This should have given me an unbiased sampling of whatever types of blogs are members of the EC network. Here's what I found:

First set:
A. Blogs served- 100
B. EC ads- 39 (39% or 44%- see conclusions)
C. no widget- 22 (one could reasonably expect that a small number of blogs would be so new to EC that they have been approved but not yet added a widget, but certainly not 22 of them. Of the 22, 2 were simply non-existent, 404 file not found errors.)

Second set:
A. Blogs served- 100
B. EC ads- 38 (38% or 43%- see conclusions)
C. no widget- 22 again (see comments above. One blog was by invitation only, one had a suspension notice. One other blog actually had a widget but I reported it for being over MY pornography threshold!)

Conclusion- if you say that I only viewed 88 (100-22) blogs in each set then the percent of EC ads is 39/88 = 44% or 38/88 = 43%! If I had discounted the 22 blogs that served no ads, for whatever reason, and continued to load blogs by this method to get 22 more valid views, even if all 22 had no EC ads, the ads would still have been 39% and 38%. Since it is more likely that there would have been some EC ads, the percentage would have gone even higher. For example, if in those 22 additional blogs, 5 had EC ads, then the percent would have been 44/100 = 44%.

For this level of ads, which is 2.5 times higher than they claim, there would need to be 500 blogs loaded with no EC ads at all to bring the percent back down to 15. Statistics is all about probabilities, and although this is possible, it is highly unlikely! Is there some little known way to load blogs whereby few or no ads are served? Perhaps I should try more sets with Favorites and Most Expensive?

To look at this another way, of the last 200 blogs I tried to open, only 156 of them had an EC widget on which to drop. Of the 156, only 79 actually had ads from other blogs in the system. That is 51%. If EC members are only shown ads from other blogs half the time (admittedly by this one method of blog delivery), no wonder they are upset.

And this is all on top of the fact that there are currently 10 paid ads in the system. I simply counted them as valid widgets that don't go to EC, but still reduce the amount of control that a blogger has over the ads which appear on his or her blog. One can choose to have up to 5 of those not display, but to remove them completely, you have to pay $48 a year and subscribe to Entrecard. This seems pretty far removed from the means of "free" blog promotion that Entrecard touts to potential members.

Checking up on the Entrecard 15% Ad Placement

web ad income Nov 24 (4 blogs, 2 web sites, writing):
Adsense $.05
Adgitize $.72
Project Wonderful $.02
Chitika $.03
Clixsense $.03
Microworkers $1.00
Swagbucks $10.00
Neobux $.065
Total: $11.92

web ad income Nov 25 (4 blogs, 2 web sites, writing):
Adsense $.03
Adgitize $.72
Project Wonderful $.02
Clixsense $.05
Neobux $.07
Total: $.89


John | Retro Programming said...

I hope Entrecard aren't cheating bloggers. It certainly wouldn't be in their best interests.

If you want to check the result over a few thousand widget impressions you could just load the ad, e.g.

It'd be interesting to see the % served if you're logged out.

It certainly looks suspicious, but I don't understand why Entrecard would cheat. They don't care about upsetting the community so if they wanted to take 50% why not just tell us?

John | Make Nothing Online said...

I forgot to say the Entrecard widget doesn't show on some blogs using certain browsers. Even the Entrecard Blog had this problem at one point with IE6. I assume it's due to the different ways browsers handle bad HTML.

Sharkbytes (TM) said...

John- I think I should check it when logged out, but I hate to because it will take all that time to load 100 pages, and I can't even get anything out of it except a blog post!

And, agreed, I should check just reloading one site 100 times.

What a project!

RE - Entrepod said...

I've done your experiment and you are correct. basically paid ads show up 5 times as much as ec paid ads.
this is why I will be getting out.
the dead widgets are another problem that they don't address. I've gotten so many penalties lately for just dropping that I'm fed up.

trust me, entrecard is going down.

Martin@Cornyman's Money-Blog said...

It was just a one-time surge of sponsored ads in the last week. Like i wrote in my Facts about the sponsored ads, it's way too expensive.

I don't believe there will be anymore so many ads at once.

First of all, the members who advertise loose themselves clicks when there are more than 6 sponsored ads.

And if you advertise only in one category, it will get boring for many members as these ads can last up to one month. Don't know if you would click each day on the same ad from the same blog.

I guess they had to raise the % otherwise there wouldn't be a lot of clicks/day for all 12 advertisers. And i think EntreCard wanted that they spent again 25$ for the next campaign, so it's better to show the ads in a faster pace :(

But they are wrong as Adgitize is much more effective even when you get just 50% of the daily clicks but for this is more than 75% cheaper advertisement!

Sharkbytes (TM) said...

Corneyman- when I did this I did not even realize that the sponsored ads had begun. I was counting a widget at that led directly to EC. Only one of the sponsored ads does this. Unless that one ad was showing up 23 or 24 out of the 38 or 39 that I counted, this is still very out of whack with the 15% commitment.

The number of sponsored ads in the system should have no effect on the promise of taking "no more than 15% of your ad impressions."