Twittering Money Away: API Service Review of Twitter Counter
Under the headline from their “Featured” web page, Twitter Counter boldly states “Become a featured user and gain more followers.”
An interesting proposition to be certain, especially as I am always looking to expand the size and reach of my collective social network which at this time includes the PI Window on Business and Procurement Insights (and Essential Connections) Blogs, the PI Window on Business Radio Show on Blog Talk Radio and, the PI Inquisitive Eye Internet TV Channel through both USTREAM and YouTube. (Note: to get an idea of how they all work together re cross pollinate, visit a recent Event Page for a conference we covered at the end of April in Washington, D.C.)
Add into the equation Twitter Counter’s offer of 15,000 views for $69 US, which is a relatively small investment, I figured it at least warranted a try.
With the initial 15,000 view, I was reasonably pleased with the result of an increase in approximately 300 to 400 followers which is about a 2 1/2% conversion rate or 1 follower for every 40 views.
What is important to note is that I checked 1 out of every 10 new follower profiles to get a feel for who they were and if they were in fact individuals that would have at least a modicum of interest in the subject matter that I cover through the network. Generally speaking the “quality” (if this is the right word), of the followers were good as they all came from a background that would at least indicate a potential for shared interests.
Based on the above, I then decided to do a second round of 15,000 views at the $69 level. This is where the promising results of the first go around, as well as the corresponding positive assessment, were replaced by disappointment and even frustration. I will let my e-mail to one of Twitter Counter’s founders, Boris Veldhuijzen van Zanten, elaborate further:
I am an author and host of the PI Window on Business Show on the Blog Talk Radio Network and I have been researching Twitter and Twitter-related tools for an upcoming article/segment.
On the two occasions in which I have used your service, the first produced approximately 300 to 400 new followers. Generally speaking, the quality of follower was reasonable good. Good enough in fact, that I decided to test your service again.
The second and most recent time, I had close to 700 follows of which the vast majority were as described in the following advanced excerpt of the post that will be the lead into the live broadcast:
“One of Twitter’s biggest problems . . . I am researching services such as Twitter Counter’s Feature, and in this second round of testing found that a great majority of those who are following are the Ruby Bells and Tia Marias and Angelina Jollies . . . I think you get my drift . . . get in the game Twitter Counter it is not the number of followers but the quality of followers that count!”
This is hardly the kind of results that would substantiate further investment, if any at all in your service.
I wanted to give you the opportunity to respond, which I would be happy to include in the article.
To Mr. Veldhuijzen van Zanten’s credit he did respond and, responded promptly:
Thanks for the chance to reply. Twitter is currently growing with 300,000 new members a day (http://thenextweb.com/socialmedia/2010/05/12/twitter-growing-by-300000-new-users-a-day/) and with that they also attract the less honest users. We do try to filter out those bogus users but can’t beat Twitter at its own game. Even if you discount the fake followers you will see that our views/followers ration is still one of the best in the industry.
In response to my subsequent e-mail requesting data relative to the follower to view ratio, Mr. Veldhuijzen van Zanten replied:
I do know that we predict 1 follower for every 100 views at TwitterCounter but often deliver 50% more than that. Not sure which package you paid for when you signed up and got 700 followers?
I then of course wrote back indicating that I was on the 15,000 views for $69 plan. But here’s the thing, when I checked out 1 out of every 10 new follower profiles with the second round, as well as a third round of 15,000 which is just concluding today, the results were markedly different. What I then did was begin checking 1 of every 3 follower profiles, as well as saving them in a designated file so that I could refer to an actual number (re the 700).
Here is a sample of the Twitter Counter new “follower” pictures from the second and third rounds:
The above “profile pictures” are of a tamer variety, but I think provide enough exposure so to speak to effectively illustrate my point.
One other trend that I found noteworthy is that in almost every instance, the later round profiles had either a paltry or zero number of followers, virtually no Tweets and had on average between 150 to 300 following them. This is an important statistic as some of the profiles that had suspiciously looking “innocent” shots demonstrated the same tendency.
There are obviously more questions such as why the first round with a relatively smaller number of follows offered a better caliber of contact, while rounds two and three had substantially more follows of which 80 to 90% were what I would classify as being undesirables.
The fact that Mr. Veldhuijzen van Zanten deflected responsibility to Twitter when he wrote that they “can’t beat Twitter at its own game,” is of little consolation as ultimately his company is the one that is receiving (or at least did receive) the $69 for the services that Twitter Counter is rendering.
In the end unless they, being Twitter Counter and Twitter, can find a way to address this problem the future of Twitter APIs which are the branches of the Twitter tree looks pretty bleak.
I will keep you posted on any future developments.