Title: Common Split Testing Mistakes 

Hello,

I’ve heard from some people that they’ve tried split 
testing before, and never really got any results from it.

But when I dig deeper and ask them exactly what they did
to test various options on their site, I kept hearing about
the same few mistakes over and over.

Making mistakes is frustrating, but it’s better to learn
from them and keep testing than to give up and dismiss 
split testing as useless.

Let’s take a look at the most common mistakes people make
when running split tests.

Stopping Too Soon

Some split testing tools advise you to end your test with 
as few as 100 visitors, but that is far too low. It’s 
advisable to run each round of testing to at least 1,000
unique visitors in order to get statistically accurate
results.

Of course, the more traffic you have, the higher the 
number of visitors it might take to get an accurate result.

Not Testing For A Full Week

If you’ve been in business long, you undoubtedly know that
almost all businesses have fluctuations in visitors and
sales on different days of the week. 

For this reason,it’s a good idea to let your tests run 
for a full week to ensure you aren’t seeing a difference 
in conversions simply because you ran one test during a 
strong sales period and one test during a weak one.

Having Too Little Traffic

In order to get a accurate split test results, you need
to have at least halfway decent traffic, or you’ll
never get accurate results.

If you aren’t getting at least 1,000 visitors per week,
you will have a hard time getting truly accurate results.

In this case, you could run some quality paid traffic to 
to your page during the test period.

Expecting Too Much

If you expect huge gains with every test, you’ll end up
being disappointed. Most tests will yield only a 1% to 5%
increase in conversions, but even a 1% gain is nothing to 
turn your nose up at over time. That 1% could be a 

The lower the traffic you have, the larger the difference
you’ll need to see real results, but any gain is better than
nothing as long as you’ve got a statistically significant 
result.

Sure, you’ll sometimes see huge jumps in conversions, but
if you expect that with every single test, you’ll be 
disappointed.

Tomorrow we’re going to talk about multivariate testing,
which is a more advanced form of split testing. 

Until then, 

{YOUR NAME HERE}

PS- Multivariate testing isn’t appropriate for every 
site. Find out tomorrow if it’s right for yours!


