A/B Testing for galleries (or whatever?)
 

I have been trying to do A/B testing try to improve click throughs on my galleries by buying traffic from trafficholder and pumping it through different versions of the same gallery like:

http://www.toosxxx4.com/galleries/sw...heHun_exp.html

and

http://www.toosxxx4.com/galleries/sw...heHun_gg1.html

The testing is starting to run into money and it is not clear that I am able to show consistently better results with one gallery over another. I haven't read about this kind of testing on the boards before. Are other people doing it? Does anyone have suggestions about doing this kind of testing cheaper or better?

Thanks,
fg

Why buy traffic? Submit your galleries and measure the results.

Isn't trafficholder blind click traffic? That's even worse than TGP traffic.

I assume you mean submit the galleries to a TGP.

I know that the Hun and worldsex ask submitters not to change their gallery after it's been submitted. Are other TGPs more tolerant? Also it seems to take somewhere between 50-300 views to get a click depending on circumstances. Submitting to a TGP doesn't seem practical for such testing.

Thanks,
fg

Yeah, where else would you submit a TGP gallery?

Why change a submitted gallery? Make your gallery changes and submit it for the next time. Make changes each time if you wish. Measure the results.

Isn't the point of this testing to submit good converting galleries to TGPs? How could they not be practical?

A good listing on Hun or Worldsex alone should send up to 50k hits in the first 30 days. Submit galleries to multiple TGPs every day and there should be plenty of TGP traffic.

I'd also say 50-300 views is on the very low end. Look for a 2-5% CTR for TGP galleries.

We seem to have different ideas about testing but I'm not sure. I want to take a gallery for a sponsor and tinker with layout and copy text and link placement / behavior for the same set of promotional pics to try to optimize click throughs for the same set of promotional pics and thumbs. If you change both layout and template as well as promotional material and sponsor, how do you know what changes are responsible for the change in performance. Is your performance better/worse because of the material or the template technology?

I am assuming you can't resubmit the same promotional material multiple times to a TGP with different layouts. Correct?

If you look at the two gallery links in my original posting you should be able to see what I am talking about. Can you submit the galleries for both links to the same TGP? I would think not.

fg

tgps dont like identical content from the same submitter...

dont overthink a tgp gallery... its better get a gallery online right now, than to stress about making the perfect gallery...its a numbers game...submit as many as you can as often as you are allowed..

First HowlingWulf tells me that A/B testing is free - just submit to tgps. Then Shemp admits you can't really do that but such testing is "overthinking".

I suppose that they know this business much better than I do and could go ahead and try to experiment as they suggest. I find, though, that I do not have such an easy time getting good sponsor content that I can work with. I spend a fair amount of time doctoring the photos and trying to leverage a few years of web development skills to try to build nice templates. I do this part time and have other projects. Overall it makes sense to me personally to try to optimize my ROI with the content. I also wonder a bit why others with more experience seem less interested in optimizing ROI for both content and the whole advertising product, if you will.

Anyway, with all due respect to the experience of HowlingWulf and Shemp, I am less interested in why I don't need to test gallery designs than with ideas for better ways to do so. Actually JollyHumper seemed to be on to a similar idea with his thread Analytics vs Clicky.

So if you are not tired of this thread and have ideas about

1) Gallery design testing.
2) Optimizing ROI for sponsor content or reasons that isn't worth the trouble

then I still await your further enlightenment.

fg

you cant A/B test the identical content on the same tgp...

I just think this level of detail is stopping you from submitting every day and making money.

Out of the 80 or so galleries I submit each month, I use a variety of templates and sponsors. Some more than others, but each one with different content. Over time, I have an idea of the CTR (click through ratio) of each template regardless of the content. I can change something, and measure the new CTR for that template - and I do from time to time. The range is fairly consistent so you're not going to discover any magic template that shoots sales into the stratosphere for TGP traffic with TGP rules.

But these changes and measuring results are far far far less important than submitting every single day.

But hey do what you want.

Listen to Shemp and Howling, they they know the TGP game from both sides.