Greetings all,

I'm a newbie to this forum. I didn't even know Michel had another forum ofther than the one at the copy doctor. I joined "The Copy Doctor" last month and one little tweak from one of those amazing videos boosted one of my client's conversion rate 25%.

Although I've been a writer for 7 years (check that, make it 8 years), writing direct-response copy is still a little "new" to me. With what I've done so far, though, I'm seeing some good results for my clients.

These days, I'm looking for the best methods for testing my own offline and online copy. Any suggestions? I've heard of multi-track generator from the copy doctor forum, but haven't tried it as of yet.

But I definitely want something for testing out my new website. Testing and feedback from other copywriters is the best way I think to improve my copy and get better results for my clients.

I look forward to learning and sharing with all of you.

Regards,
Carlon

Carlon,

Welcome!

MTG is the best around.

However, I often recommend that someone new to testing MASTER A/B Split testing, before jumping into multi-variate testing.

By relying totally on multi-variate testing such as what MTG does, you will miss some of the important (and more subtle) lessons that your need to learn about testing.

Such as developing a true understanding for how to determine (and understand) statistical significance.

Or how to develop a real appreciation for the impact that a single, small change (such as one word in your headline) can have on response rates.

Too often when we are testing half a dozen variables at the same time, we just look at the improvement as a whole and fail to ponder the possible reason a specific change produced a better (or worse) result.

It is in the pondering, that we can discover "principles" that can be applied to other aspects of our testing.

Multivariate testing is powerful. Very powerful. But just like teaching someone to fire a gun, it is best to start out with a single shot rifle and master the finer basics of marksmanship, before grabbing a machine gun and blasting away.

Happy Testing!

Eric

Good response Eric,

I agree. Multi-variate testing can be very powerful, but it can only be as good as the copy fed into it, the strategy used, knowing which elements to test, traffic quality, and more...

For example:

It's no small task to write several potential test elements, such as various headlines, sub-heads, etc (that will be mixed and matched randomly), so that they all work with each other in a FLOW no matter what combos the multi-variate dishes out to the readers.

It's not a good idea to just bang up several headlines if some of them make no sense with the rest of the copy - at least if we want to use the tool to its max. potential.

Let me take a step backwards. The way I see multi-variate testing, is like adding a turbo-booster to your car. Before you add it on, you have to make sure your tires, suspension, etc, can handle the new motor. Because it can only be as fast as your weakest link.

In other words, the best use of the multi-variate testing tool, IMO, comes after the sales copy is already a solid work, with good sales already happening.

i.e.... very good copy, a proven marketing strategy, good quality traffic flow already in place, etc. It's a tool that allows experts to beat solid "die-hard" controls much more quickly.

Just like critiques, it is easier to critique a good piece of copy, than it is to critique copy that is flawed throughout the whole piece - better to toss it and start over until all the elements are solid, rather than "band-aid" it to death.

Using multi-variate testing on a poor piece of copy, is like entering it into a crutch race... you can only hobble so fast. It's not a beginners tool, and will never substitute the basic "sprinting speed" of a highly trained copywriter/conversion expert.

Carlon, focus on the basics, and use multi-variate testing to "ice the cake" when you feel you've hit a wall with a piece that is already performing.

All the best,

Exactly Tim!

That is one of the biggest mistakes I see marketers make with multivariate testing. They lose site of the fact that all of their test variables need to work together regardless of the way they are combined.

In my opinion this is the major drawback of pure Taguchi based testing. The algorithms are not designed to be predictive of the effect that combinations of variables can have on the overall conversion rate of the sales copy. This is where MTG is much better than pure Taguchi based systems.

But in the end, the value of any testing tool is limited by the quality of the variables you feed into it.

Eric

Eric and Tim,

Your advice and words are much appreciated. I was pretty unsure how to proceed with testing in such a way to get the most out of it. But now, you have given me an idea of where to start.

I will try out A/B split testing first and then go from there.

Of course, I'll first be sure that the copy is getting results.

Thanks again,

Carlon