The Dirty Little Secret Nobody Tells You About Website Split Testing...

I've got a confession to make. This may come as aversion won. I've conducted price tests in which we
surprise to you (after all, I'm a testing fanatic)... But,tested $17 vs. $97 and $97 converted better! (Not
testing doesn't improve website conversion everyjust higher visitor value, but actually higher
time.conversion!)
Nearly all "online marketing gurus" promote theSo it really is impossible to know in advance what
benefits of conducting split tests and multi-variableversion or variable is going to win. Experience and
tests on your website. But, few (if any) tell you theprior testing data can help you make an educated
WHOLE truth about how to test, what to test andguess, but you still have to test it and see.
what results to realistically expect from your testing.I view selecting test variables, kind of like picking
The consequences? Many online markers becomestocks. You can do all of your homework, use a
confused by and disappointed with their first fledglingvariable that has boosted response on every other
attempts at conducting testing campaigns on theirsite you tested and still not have a winner.
websites.And the opposite can happen. You can test a variable
Because their results often don't live up to thethat you are almost positive will be a dog, (or even
"hype", they become discourage and simply assumemake an error when setting up your variables) and it
that "testing doesn't work" or "it's not worth thecan boost response. I've often seen headlines that I
effort".hated or graphic images that were down right UGLY
I hear this from my clients all the time. They will tellconvert better than the more appealing versions.
me, "oh... I tried testing on my site and it didn't help."So just like picking a stock, past performance doesn't
Well... in an effort to clear up a few of theguarantee future results! Every single variable you
misconceptions about testing and help folkstest will not be a winner. In fact when it comes to
overcome some of the unrealistic expectations thatthe stock market, if you can pick a winning trade
simply set them up for disappointment, I'm about tojust 51% of the time, you'll become rich.
tell you the "dirty little secret" that the gurus usuallyThe same principle applies to test variable selection. If
fail to mention.you're able to consistently pick winning variables 50%
Every variable you test will NOT automaticallyor the time or more, you're doing an outstanding job!
improve sales. In fact MOST of the variables you(Or you're simply not running enough tests...)
test will either reduce response or make noNow, am I telling you all of this to discourage you
measurable difference!from testing? ABSOLUTELY NOT!
In fact, after conducting over 6,000 split andLike I said at the beginning of the article, I'm a
multivariable tests over the past six years I'vetesting fanatic.
generated some pretty good data on what you canI've seen small businesses transformed into "big"
expect to achieve in a typical testing campaign.businesses almost overnight via proper testing.
In most multivariable tests only 1/4 to 1/3 of yourI've seen a laid off, single father go from struggling
variables will boost response. Another 1/4 to 1/3 ofto pay the bills selling an ebook to making a solid six
your test variables will actually reduce response. Andfigure income by only tweaking and testing 4 items
1/3 to 1/2 of your variables won't make aon his sales letter.
measurable difference.Done right (and with proper expectations) testing can
(Well... that's not entirely true. If you run any testindeed change your life and transform your business.
long enough a winner will usually emerge, but it is notI just want to make sure that you go into the
worth the trouble to let a campaign run for monthsprocess with the right, long term mindset and that
on end just for a 1/10th of a percent improvement inyou don't become discouraged and quit if your first
conversion.)attempts at testing fail to yield earth shattering
Of course, the more experience you gain in testing,improvements in response.
the better your odds of finding a variable that will beThe good news for testing is, because you can flush
a winner... but you still have no guarantees.the variables that don't boost response and only
Often variables that you are sure will boost sales,keep the "winners", even if you only pick a winner
wind up having no effect or worse, decrease10% of the time you'll still make huge gains in
response.conversion and response over the long haul.
I've seen tests where a 1 year guarantee wasHappy testing!
tested against NO guarantee and the no guarantee