FREE REPORT (if you qualify): How
and why you
should split test - even with just 10 visitors a day...
You may have heard marketers
talk about the importance of "split
testing", but perhaps you're thinking...
(a) I don't have enough traffic, or
(b) I don't really know how to do it.
The truth is... even if you have just 10 visitors a day, you can and
In fact, you're
losing potential money and stunting your site's potential for growth if
Even if you just have 10 visitors a day, that's about 300 a month.
Let's say you convert just 1 in every 100 visitors into customers (i.e.
a 1% conversion rate), then at 300 visitors a month you make 3 sales.
To double your sales, you could double your traffic - a no-brainer,
right? That means finding double
the sources of traffic.
You could raise your conversion rate from 1% to 2%...
That means 300 visitors a month produces not 3 but 6
sales... i.e. this
double your sales.
In other words, you're getting more out of the traffic you've already
got. And you could use the extra money generated by these sales to...
get more traffic!
My report, "How To Split Test With 10 Visitors A Day" explains exactly
how to raise your conversion rate through split testing, even if you
have just 10 visitors a day.
To qualify for a FREE copy of this report, here's all I ask you do:
Read the remainder of this sales letter for my split testing
Answer the one
simple question posed near the bottom.
OK, given that you can
split test even with just 10 visitors a day, why would anyone pay for a
split testing tool?...
Why not just use a free split
When people first "get into" split testing, they are often advised to
get a free script to do it - after all, there are a few free split
testing (and even one or two multivariate)
"tools" out there. The problem is... this advice could be bad
There are two reasons. The first is important, but the second (in my
opinion) is critical
The first issue is...
As with anything in life, there's usually a reason
free. Some of the "free" scripts I've tried come with literally no
support, so if
you get stuck at any stage, you're on your own. And let's face it, if
you're new to split testing, the last thing you want to be doing is
struggling to come to grips with the script!
Other "free" scripts I've encountered are designed to be "upsells" -
just about good enough to get by, but with enough weaknesses to get you
to want to "upgrade" to an expensive version.
(Or in the case of
Google's testing tool, the upsell is Adwords. And your conversion data
useful to them...
after all, they also set your Adwords prices, so I'm sure they'd love
to know how
much money you're making from a testing campaign. Think about it...!)
However, important though that is, there's a second (often overlooked)
issue which is much less
familiar to people who are new (and even ones who are not
It's absolutely critical
that the split testing tool you choose is accurate
Otherwise, your data could be as good as worthless!
Before I show you why
it could be worthless, let me share with you a little secret about how
split testing scripts work...
What happens when the "cookie"
Many split testing scripts rely
solely on "cookies", little bits of information that are sent to the
browser. The problem is, cookies are not entirely reliable - they can
be deleted or blocked from a
visitor's browser. When
that happens, those scripts can't track visitors any more, making them
combine cookies with IP tracking, i.e. they record the visitor's IP
address as well, so that if one fails, the other acts as a "backup".
This is a definite improvement over using cookies
However, a visitor's IP address can change each time they
connect to the Internet! So IP tracking is only reliable in the short term
And yet - most testing tools on the market rely solely on these methods!
methods of tracking - why
that's critically important to you...
offers an optional third
method of tracking visitors, called the "unique tag identifier" (or
UTI). It is most
effective when combined with an email subscription. A subscriber clicks
your link in an email message. If they've been to a campaign page
will show them the correct version -
the dependency on cookies and IP address
important of all... if the subscriber takes the action you desire,
the action is attributed to the correct version, ensuring
Which begs the following question, doesn't it?...
are split testing stats, really? An insider spills the beans...
about it. The job of a split testing tool is to assign an action (such
as a sale) to the correct version the visitor originally saw when possible
After all, you're basing your decisions on these results.
because most scripts rely solely on cookies and IP tracking (which are
not permanent), how reliable are the stats you're getting?
And yet, it gets worse...
simple fact that not
can be tracked long term - so if a
visitor deletes their cookies, and takes action on your site from a
different IP address, what do most split testing scripts do with that
As an insider to this industry (I've also been using split testing
scripts since 1998), let me tell you the answer...
Most just guess!
they assign the current version to the visitor.
In other words, if the script can't work out which version the visitor
originally saw, they will just assign it to the next version due to be
rotated... effectively killing the accuracy of your stats!
Even worse, if the unidentified visitor (i.e. someone without the
cookie or IP address matching) comes to an action page again
most of these will count the action again!
Why is this so important? Because...
... even a single
incorrect action assigned to the wrong version could have a critical
effect on your decision - let alone 2 or 3!
Let me give you a quick example:
How your split testing stats can
be "killed" by just two bad apples
instance, let's say you had 12 sales from your current sales letter
(which we'll call Version 1), and 18 sales from
a variation you're testing (call it Version 2). Based on these figures alone
, you could be
about 85% confident
that Version 2 is an actual improvement, and not down to chance. (I
know this sounds obvious, but you'll see in 2 minutes why it's not the
what if your split testing software had really been "guessing" with one
or two of those sales? What if one of them was really a randomly
assigned duplicate visit, or just incorrectly assigned?
If just one
of those actions was assigned to the
wrong version (i.e. the split was really
13/17), you can be much less confident the difference was not down to
chance. And just one
(at 14/16), and there's really no significant
difference at all! (The difference is quite likely to be down to
words, it's absolutely
to get accurate data!
This is why Power
Split Tester™ never
after attempting its 3 methods of identification - the "unique tag
identifier", cookies and IP address - if it can't work out
which version the visitor originally saw, it is brutally honest and
will classify the action as "unattributed".
This is not
a failure - quite the opposite, because it's a fact
that not all
visitors can be tracked in
the long term
. So any
split testing program that magically assigns all
visitors to an action every time without fail is inevitably
some kind of voodoo magic
system, almost certainly called "guessing".
Instead, when Power
can't determine the source of an
action, it enables you to compare the logs (see below), to determine
whether these are real actions (but ones that couldn't be tracked
automatically), or just repeat page visits (bogus actions) - and you
then assign (or delete) these actions yourself.
In other words, the
combination of not just 2 but 3
tracking methods, as well as not
guessing when it
can't identify the true version, means that (in my opinion), Power Split Tester™
is quite probably the
most accurate split testing tool available
(If you don't
me, ask my "competitors" how many tracking methods they use, and what
their program does with actions it can't trace! I think you'll be
shocked by the answers you get - if you get
a straight answer. Remember, it's impossible
actions to be automatically traced - cookies and IP address are not
How to have real
confidence in your stats:
I mentioned earlier, a 12/18 split between two versions of a sales page
means you can be fairly confident the second is an improvement over
However, at what point does the element of chance come in? For
instance, is 13/17 a real improvement? What about 14/16? Both seem
improvements, but mathematically speaking, it's quite likely the
difference is just down to chance.
don't worry... you don't
have to be a math genius to use Power Split Tester™
program will tell you just how confident you can be in
Just remember, you're always trying to beat your
original version, which in Power Split Tester™
is always Version 1. In
the picture below, Version 2 is ahead, and the program is reporting
that you can be about "95% confident" it's an improvement over
original, Version 1.
"95% confidence" basically means there's a 5% chance you could be
wrong, i.e. that Version 2 isn't really
an improvement, it's just ahead
by chance. Power
will report at approximately 85%, 90%, 95%
Hint: Some suggest
you should aim for a 99% confidence... but that might mean you need a lot of
data. In the end, Power
can tell you how your stats
are doing, and it can give you recommendations based on that data.
But you're the
it's up to you
to make the judgment call about what to do with that information!
Easy-to-read logs help you to
make better decisions
web site has a log file, but Power Split Tester™
keeps its own log
files, and logs every visitor to a campaign and action page, and you
can view these in an easy-to-read format, right from your Admin Panel.
This allows you to quickly see other information which a standard split
tool wouldn't tell you, such as where
these visitors are coming from - and which "version" they saw, which a
standard log file wouldn't tell you.
can help you make decisions about your campaigns that raw stats alone
wouldn't be able to tell you. Plus, you can use these log files to help
you determine if the "unattributed" actions are real
Unlimited campaigns and split
, you can run as many split testing campaigns
as you like, and you can test as many variations as you like!
a simple split test with two versions is called an A/B test, but you
can also run an A/B/C test (3 versions), an A/B/C/D test (4 versions)
and so on.
Track up to 10 actions per
Most other split testing tools only allow you to track one particular
action. However, with Power
you can track up to 10 different actions
enables you to easily
monitor what your visitors are doing
example, if you offer an email subscription, you could track both requested
Or if your sales letter is broken up into several pages, you could
track click-thrus to each page.
Hint: Action 1 is
always treated as the primary action, upon which Power
its advice, so you'll want to set action 1 as
the principal goal of your campaign page, such as a subscription, or
In summary, here's the powerful
unique) features of Power Split Tester™...
- Up to 3
different methods of tracking (cookies, IP address and
"unique tag identifier") enables vitally accurate tracking.
(Remember, cookies and IP
address are not
guaranteed methods of tracking over the long term.)
- Ability to combine the "UTI" with an email
subscription to ensure
your email subscriber sees the correct version, and the
subscriber's action gets attributed to the correct version, even if
cookie and IP tracking fails!
and easy to read logs available to you from the Admin
Panel enable you to see what your visitors
are doing, and which version of your campaign they originally saw.
Also, helps you
to identify genuine and bogus actions.
"guessing". When a visitor takes an action you're
tracking (such as a subscription or sale), if Power
Split Tester™ cannot work out which version your visitor
originally saw, the action is classed as "unattributed". However, you
can use the logs to manually assign it to an action. This ensures there
guesses (after all, cookies and IP tracking can never be perfect),
bogus repeat actions skewing your data - absolutely essential when
you're making decisions based on your campaign results!
campaigns. There is no limit to the number of split
testing campaigns you can run on your site, even at the same
remember you'll need the traffic to handle them. If you're only getting
10 visitors a day, stick to one simple A/B test for now.)
versions. You don't have to limit yourself to a simple A/B
test (with 2 versions). If you have the traffic, you can have as many
as you want in a campaign.
different elements of your site. If you need to rotate
parts of your site that are far away from each other, you
can do that easily. So it's no problem to
rotate a matching header and footer, or to rotate the price or some
other element throughout your entire
up to 10 actions per campaign. This enables you to track
many different actions in a single campaign - for example: requested
subscriptions, and sales.
This can help you to quickly spot weaknesses in your sales process -
perhaps you have a version where people are not confirming their
subscriptions as much. This will probably have a knock-on effect on
your sales! Yet you wouldn't know this without tracking both requested and confirmed
each version a "weight". If you already have an effective
sales letter, you may not want to conduct a split test on all your traffic.
You can give each version a "weight", which tells Power Split Tester™
of your visitors should see each version. For example, if you only want
to test on
a small proportion of your traffic, then you could tell Power Split Tester™
to send say 90% of your visitors to your current
version, and just 10% to your test variation. Or you could just do a
straightforward split test, with 50% going to one version, and 50%
going to another - it's entirely up to you!
- Be confident
in your data. The "stats" facility will report when a
version is an improvement over your original, at "confidence levels" of
85%, 90%, 95% and 99%. This tells you how confident you can be that the
version is genuinely
an improvement, and not down to chance.
advice. The "stats" facility doesn't just report the
cold, hard numbers - but gives you suggestions
as to what they actually mean and what you might wish to do at a
particular stage of testing.
- All the
data is "for your eyes only". Since Power Split Tester™
is installed on your site, you
never have to give a third-party your valuable conversion data. (I'm
simply amazed by people who are happy to give away their
conversion data to Google - the same company that often sets their
Adwords pay-per-click prices! Think about it...)
- Easy to
install and use. Despite all these features, I've
worked hard to make Power
Split Tester™ as easy to install and use as possible. The
last thing I want you to be doing is struggling to get it working! You
need to mess around with a MySql database. For that reason, the average
installation time is just 6 minutes (if the auto-installation script is
used), and you'll be running your first test campaign after about
another 10-15 minutes!
|"Review of Paul Hancox's Power
The truth is that the first time I looked at the Power Split Tester
Instructions I thought that the install and use would be complicated.
Well, I couldn't have been more incorrect. Paul gives you detailed,
step-by-step installation instructions that will have you up and
running in a jiffy. My install took all of 5 minutes.
Power Split Tester even comes with some sample files so you can quickly
and easily set up a test campaign. Setting up a test campaign couldn't
have been easier. It only took a few minutes.
Once I saw how easy it was to set up the test campaign I immediately
created a real campaign. 10 minutes later I had a split test campaign
setup to see which headline would get me the greatest number of
subscribers to my newsletters. Right off the bat I can see that my new
headline is pulling in twice as many opt-ins as my old headline! What
would a twofold increase in newsletter subscribers do for your business?
While it's always important to get more traffic to your website I now
realize that it's more important to make the most out of the traffic
that you already have. Power Split Tester is a tool that can help you
I recommend Power Split Tester because it is easy to install, easy to
use, and does exactly what it says it does.
Tony Spann (Warrior ID: "TeamGlobal")
I use Paul's Power Split Tester and it is outstanding! You
can test everything with it from the headline to the whole
sales page plus much more.
I used another one of Paul's products for years and it was
outstanding too so I knew this one would be great. It's like
I told Paul when I wrote him...."When it comes to split
testing Paul is King". He really knows his stuff.
I'm a true believer in testing...if you're not testing your
sales page than you could be losing thousands or even tens
of thousands of dollars or more.
Keep up the good work Paul! Absolutely amazing!
Charles E. White (same Warrior ID)
Start testing today,
watch your conversion rate soar...
is a program that you install on your
requires PHP 4.3 or higher, which is available on virtually all web
hosts. (A database is not
Full instructions on how to install and use it are provided, along with
90 days of technical support. You may use it on an unlimited number of
domains that you own, but it may not
be re-distributed or sold to third parties.
are also protected by my 90 day personal guarantee. If you find the
program isn't for you within 90 days of purchase, then email me for a
full refund. (In this case you must remove the program from your
If you have any questions before your purchase (or even after!), don't
hesitate to send me an email. Just use the "Contact" link at the
bottom. (You could check the Frequently
page first to see if your question has
already been answered.)