Comments on: A/B Split Testing Versus ABAB Split Testing

By: used tires

used tires — Tue, 13 Apr 2010 20:37:22 +0000

Haha, I thought the same too! Showing my age here, lol.

Till then,

Jean

By: body building supplement

body building supplement — Thu, 01 Apr 2010 06:13:32 +0000

Agree with your view that “In fact,ABAB method is a common way of split testing in other fields like Pyschology. These fields use the ABAB method for most experiments”. I will consider your recommendation to use the ABAB method.Thanks for sharing great blog.Worth to read your blog and appreciate it.

By: Aluminum Case

Aluminum Case — Mon, 22 Mar 2010 22:51:55 +0000

Thanks for the advice Mark. I wouldn’t say our conversion rate is horrible. I just find it tough to test things when the conversion rate is below 2-3%. Sometimes we have to look at factors other than just the sales page design. In our case we offer a niche product but we get a lot of traffic that is looking for slightly different products. I just know that we would have to test things for quite a while to get accurate results. There will always be people who won’t buy from you regardless of how good your design is and there will always be people who will buy regardless of how bad your design is. So you need a big enough data sample to dilute those cases.

By: Makeup Store

Makeup Store — Wed, 10 Mar 2010 04:23:27 +0000

I’ve only just started split testing recently, however, I have seen excellent results. It just takes time to debug and figure out the correct method to follow.

By: used tires

used tires — Tue, 09 Mar 2010 10:33:58 +0000

Yea, it seems like it’s not being looked at seriously by quite a few marketers for now but hopefully in the future, they’ll realize it’s usefulness. Just takes time for everything to catch on, I guess.

Till then,

Jean

By: Tom - marketing tips

Tom - marketing tips — Mon, 22 Feb 2010 12:48:14 +0000

Good information about split testing.I guess many marketers are not doing split testing and are not getting the results they were looking for.I will bookmark and share this with my friends.

By: Mark Acutt

Mark Acutt — Thu, 18 Feb 2010 00:22:57 +0000

Hi Aluminum Case,

I came across this post and found it interesting. There’s obviously A/B split testing (which I use often) as well as MVT (Multi Variant Split Testing – easily setup via Google Website Optimizer).

Anyway, MVT is a different ballpark as it swaps and test various different areas of your web page to find the best combination and highest conversion rate. It’s really cool and works well (even in blogs).

Regarding your issue: What do you do when your conversion rates are extremely low?

Try something totally different. Monitor what your competition are doing. Use tools (like PPC Bully) to see how long your competitors ads have been around. A good profitability indication is if the ads have been around for 30 days or more.

Nobody will continue advertising is they’re losing money — right?

Then draft a new landing page/s around what appears to be working.

Please don’t click on PPC ads, copy the URL’s and paste it into a new Browser window.

You should be able to get a 2% – 3% conversion rate. Some of my sites get 5%+.

There are many different styles/types of landing pages. Be careful or ‘review’ only page – Google frowns upon these and your QS (Quality Score) will suffer severely!

I usually use PRESELL type sites… lots of unique content targeting the visitor. Essentially I pre-qualify the visitor before they hit the vendor’s sale page (it also keeps Google happy and PPC bid prices low).

This doesn’t mean it works in ALL markets, but works very well in most.

Also, make sure you now comply with the latest FTC policies (December 2009) or you could be shut down as an affiliate.

I hope this helps.

Best,
Mark Acutt

By: Aluminum Case

Aluminum Case — Tue, 16 Feb 2010 00:03:11 +0000

Well I was hoping for a response, but I guess you’re pretty busy. I guess my only option is to use a larger data sample with a longer time period.

By: Aluminum Case

Aluminum Case — Tue, 05 Jan 2010 20:56:14 +0000

I agree with your points about split testing. For it to be effective, you need to take away outside variables that could affect things like day of the week or time of day. To get an accurate comparison, you need to give each version an equal chance.

What do you do when your conversion rates are extremely low? When you only get 1 sale out of hundreds of visitors, using different landing pages might not make much difference. The conversion rates may still come down to other factors such as price.

By: used tires

used tires — Sun, 03 Jan 2010 10:48:05 +0000

This comment reply sounds like… humm… whats the right word… a straight copy? lol wow.

Anyways its great that you decided to follow up that email question here on your blog, I got some good take away from it and I am sure many will too, and plus… now you can just link to this in case you get more emails about this lol.

Till then,

Jean