Hey Stephen

I agree with you. I'm curious though... you say, "but I rarely ever see evidence of anyone testing".

Split testing is done "behind the scenes". You're not supposed to be seeing where a person's testing

I think the Number #1 reason why the bulk of people still aren't testing is because they're not generating enough traffic and therefore sales. You can't do a particularly effective split test on 50 sales - you need hundreds to start getting good data.

However, that's not really a good excuse. I wrote a free report a while back, on how people can split test even with just 10 visitors a day (I really need to update it):

How To Split Test With 10 Visitors A Day

Why everyone isn't doing this is beyond me.

But split testing the finer elements of a sales page, i.e. 2 bonuses instead of 3, requires lots of sales, and I suspect people don't do it because they don't have the sales.

For these people, modelling success is indeed a great first step... although there are other things they can be split testing, until they start getting the sales.

Split testing is not hard (although many people believe it is) but it can be tedious...especially if you don't have a feel for what people's hot buttons are. I actually learned a lot from Karon Thackston that put split testing copy into perspective for me and finally made sense. However, I happen to agree with Benjamin here when he says that marketers tend to tire and don't want to put in the extra effort to follow through with the testing.

I certainly don't get excited about testing, even though that sure would make it easier!! And, unfortunately, still view it as a necessary evil. If I could figure out a way to make it fun, I'd be set!

I feel the same way. It's the "right" thing for people to say. I opened a thread on video sales letters for instance and I knew I would get a million and one responses saying "test test test". Obviously I know it's something I needed to test but there is no harm in finding out what other peoples experiences were.

I did end up testing video on one of my sales pages. The results were minimal. Actually, there was no change at all, after 150 conversions. So I reverted back to the original and decided to split test something else.

So yes, there are some of us out there that do test. And Paul H is correct. The reason most people don't test is because of a lack of traffic. Which is understandable. The trick? Use a different conversion metric besides sales.

Cheers,
Matt

My experiences with split testing never cease to amaze me.

What I think is great is quite often wrong; and since split testing is very easy to do (I use Google Web Optimizer), the information I get back is priceless.

Not only that, but I use what MikeHumpheys said above... Taking little test results from some of my own sales experiments allows me to craft better copy for clients; and I use those results everywhere.

Stephen asked about evidence of testing; here are a few results that I have discovered thru simple split testing...

A) Changing one sentence in the resource box of one of my EzineArticles resulted in an increase of almost 13% in my CTR to my anchor text

B) Keeping the headline text exactly the same, I tested the control (regular text) vs a Photoshop image of a different font, and saw an increase in conversions of 0.4%
...

While not a "Stop the Presses!" increase, that simple tweak generated one extra sale (roughly) for every 200 visitors that were coming to my site anyways.

3) One other thing that I tested that surprised me... I tested an Adsense block on the bottom of a sales page I was running... Where they would be seen only after the visitor had gone thru the rest of the sales page.

While this goes against every expert rule, I actually found that conversions on the sales page increased on every single page I tested.

Why? Don't know... maybe the visitor thought that Google's presence on the page added more authority or whatever.

Bottom line is this... If you don't test, you simply don't know.

And while you may not see massive conversion rate differences, a 0.2% here and another 0.4% there, and one last 0.3% over there adds up to another 1% (actually 0.9%) conversion rate from the very same visitors that are coming to your page anyways.

And if your page was only converting at 1% in the first place, you have doubled your conversion rate (and income).

Since it is very easy to do, you should be testing as often as you can; regardless of the amount of traffic you are getting.

My opinion... But it works for me.

Hey Stephen,

I'm gonna be cliche' for a minute...

I am guilty of dishing out this advice almost anytime I feel it's the right answer...

BUT...

Take Matt's example. He wanted feedback on experiences with video.

I have no idea what his product or market is. Even if I did, for all I know he could have a great sales letter and a horrible video, or vice-versa.

He could have paid a pro for the sales letter and outsourced the video for a couple of hundred dollars.

He may have wrote both and done well, but may have a monotone voice.

Maybe his headline/lead in the copy did a great job of pulling readers in, while a similar approach in video would turn visitors off.

Maybe just the way he delivers one line in the video does something to hurt credibility with the viewer.

It's impossible to know this stuff.

Even with my own stuff, I almost never guess which headline will pull the best. Or which price point will be the most profitable.

And when people are asking for advice in forums, we really don't have any idea of what the bigger picture is.

Even if someone is in the same market as me, I could say "video killed it in this niche", and they could find video bombs for them. Because we'd have different approaches to the market, and a thousand subtle yet important differences in our approaches.

So when it comes to questions like...

- Should I use red or blue for my headline?

- Will video work in my niche?

- How many bonuses is too many?

- $37 or $67?

- How many visitors can I get if I write 5 articles a day for a year?

I suspect "test" is the most common answer. Because unless we're behind the scenes and get to look at all of the little nuances of their business, we're not really in a position to give even a remotely educated guess.

Now, even if people did share results, which I personally wouldn't do in detail on a public forum, the results would still vary wildly from one person to the next.

It's kind of like swiping. Without understanding the psychology going on in what you're swiping, you won't be able to get the same results.

Without understanding why a video works well for a certain market, or a layout works well or whatever, it's hard to effectively "borrow" the reasons why it works in the first place.

Hope that makes a little sense. And explains why I defer to the old "you have to test" answer on a pretty regular basis.

Take care,

-Scott

If it wasn't for "testing" I would have gone broke a long time ago...I cannot count how many times my sales went down by 100% and after several test I bumped it BACK up 100%.

As Eric Graham says...in this business you are either growing or your dying..there ant no in-between. Actually I think Eric got that from the movie "Tommy boy."


Test Or Die! Thats what I say.

PS to be honest I dont see how any business survives online long term WITHOUT testing.

[DELETED]

So a few people said the #1 issue for people and split testing was...they are not getting enough traffic...

ok thats cool...all i have to say is "dont put the cart before the horse" if your not getting enough traffic...then dont worry about split testing...get more f-ing traffic!!!

dont waist 1 second setting up a split test (unless your buying traffic and itwill be a lot)

when im starting a new campaign (i only buy traffic) ill set up a very simple test just using aweber.

just create 2 copies of the same forms, put 1 on a page with a separate headline, then simply send 1/2 the traffic to each page. my tracking is in aweber the sign up rate on each form will show me % on each page...total time invested....maybe an extra 2 minutes?

then if i see some results that i know are close to profit, ill take the time to set up tracking 202, and really start to test...why waist time?

you can sit there all day trying to tweak your 10 visitors but it wont matter. either way your going to make little to no money b/c you HAVE NO TRAFFIC.

long story short, get traffic, then test when you have it...