Tracking And Testing
If you have not done so, it is recommended you read Part 12 – Sales Copy, and all previous tutorials before commencing this one.
Welcome to the thirteenth unit in the ‘Journey Into Internet Marketing’ tutorial series.
In this tutorial, we are going to ensure that you fully understand the twin concepts of tracking and testing how well your site is performing.
More importantly, by the end of this tutorial, you will understand why these concepts are so important and how you should apply your knowledge to your own online business.
As you will discover, the basic ‘in a nutshell’ reasoning behind both tracking and testing is that when used in combination, they enable you to improve the efficiency of what you do.
In this way, tracking and testing ensure that every action you take yields the maximum possible returns, thereby optimizing the effectiveness of your income generation efforts.
Tracking and testing is something that far too many new online marketers completely ignore, mainly because they are not concepts to which they have been introduced. On the other hand, most super successful online marketers also happen to be rabid fans of tracking and testing.
It is no coincidence that the most successful Internet marketers are avid trackers and testers to a man whilst those who struggle ignore the value of tracking and testing completely.
You are about to discover just why this is…
You have seen the tracking concept in action in previous tutorials, but as a reminder, tracking is all about knowing how your site is currently performing.
You saw an example of this in tutorial 6 when we looked at how you track the statistics of a review site and in tutorial 9 when we considered how you could use the statistics provided by AdWords in combination with those of your affiliate network site to establish your conversion figures.
The basic concept of tracking is always the same.
Once you know how well your site is achieving your objectives, you can work on improving the performance. However, if you have no metrics by which you can measure current performance levels, then it is impossible to improve them.
In addition to using propriety tracking systems such as those built into most PPC advertising and affiliate network sites, there are also many tools and resources which enable you to track the performance of your own site yourself.
For example, tutorial 6 introduced the idea of using one of the statistics programs built into the control panel of your site to track visitor numbers:
Even basic analytical tools like the stats programs inside your control panel enable you to collect together a mass of information about your visitors.
For example, scrolling down the page allows you to see the average amount of time a visitor spends on your site:
In this example, the average visit from each of the 548 people who have been to this site is 119 seconds, a single second short of 2 minutes.
Okay, imagine that you have a site which features a long copy sales letter selling your product. You know from your own experience and that of other people who have read your sales copy that it takes between five and ten minutes to do so.
In this scenario, an average visit time of 2 minutes would clearly indicate that you have a very large problem.
No-one is staying on your site long enough to read your sales letter!
From this simple example, you can immediately see that even the most basic statistical information about your site can highlight major problems.
Further down the statistics page, you have a list of the 25 most popular pages on your site, a list of sites from where your traffic is coming and the keyword phrases that are sending targeted visitors to you.
All of this information is critically important because it points out exactly where you need to tweak and finesse your current marketing system.
For instance, you may find that some of the keyword phrases that are sending traffic to your site are not keywords that you have consciously targeted. This might suggest that a change of keyword focus could improve your visitor numbers and therefore your profitability.
In short, the pre-installed statistics programs that you access through the control panel of your site are excellent for highlighting basic statistics about what your site is doing. However, my choice would always be to install the excellent free Statscounter tracker on your site as this a program that provides even more detailed information about your site performance.
If you have not used this program before, follow the link to create a free account. Once you have done so, log in to begin the site setup process.
Click on the ‘My Projects’ link at the top of the page and then ‘Add New Project’:
On the first project setup screen, add the title of your site at the top followed by the URL and the category into which your site fits best:
Set the maximum visit length at somewhere between one and six hours before scrolling to the bottom of the page to move onto the next screen.
Here, you click the large blue ‘Configure & Install Code’ arrow:
After this, you have the option of the kind of counter you want to install on your site. By default, it is set to show a visible counter when Statscounter is installed on your site.
This is definitely not something that I would recommend especially in the early days before you start attracting reasonable visitor numbers.
A counter that shows 37 or 78 visitors looks very sad but more importantly, it suggests to your visitor that your site is unpopular. This immediately plants a seed of doubt in their minds, which is the last thing you want to do.
For these reasons, I always choose an invisible counter as shown here:
Click ‘Next’ before selecting the ‘Installation Guide’ that you want to use from the drop down menu at the top of the page:
There is as you can see an option to show the ‘NVU Editor’ installation guide. But if this is not the HTML editor you used to create your site, then choose the most appropriate guide from the drop-down list or use the default guide.
The guide you choose appears on the next screen to tell you how to install the Statscounter code on your site. Consequently, if you already know how to add code to your page – if you have already installed a subscription form code as an example – you do not need to worry overly much about this.
Ignore the two optional choices beneath it and move on to the next screen:
Now you have the stats counter code and the instructions about how to install it on your site. Follow the instructions and install Statscounter.
Once the counter is installed on your site, the information is presented on the Statscounter dashboard page like this:
As you can see on the right hand side, you have a snapshot of your visitor numbers. However, notice that the icon shown on the far left indicates that the stats presented by this tool are in real time. This is what makes Statscounter more valuable as a tracking tool than the stats available from inside your control panel which are updated every 24 hours.
Statscounter on the other hand lets you see exactly what is happening with your site ‘live’. This naturally ensures that you have a far clearer picture of what is happening once your site starts getting busier.
Once you are tracking your site performance, you will soon form a clear picture of how well it is doing.
If it is doing pretty well, your tracking stats will tell you. If on the other hand it is not doing so well, you may be able to see why from your statistics. The previous example of 119 second visits on a page that takes 5 minutes to get through clearly demonstrates this.
Okay, so from this example, you know that people are not reading your sales page, a fact which will surely be reflected in your sales figures. What you don’t know is why this is happening. And of course, if you don’t know why something is happening, you can’t do anything to remedy the situation. You are losing money and will continue to do so because you don’t understand why your sales system doesn’t work.
This is why constantly testing everything you do in your online marketing activities is so fundamentally essential. Without testing, every single decision you make is based on guesswork and ‘hunches’ and whilst some (or perhaps even many) of your decisions will work out, far too many won’t.
For example, imagine that your tracking indicates that less than 1% of targeted prospects who visit your squeeze page choose to subscribe to your list. Assuming that your visitors are sufficiently well targeted, I would expect a conversion rate two or three times higher than this from even an average squeeze page. Something is definitely going wrong if your subscription rate is less than 1%.
Okay, so maybe you think it is the graphics that are letting you down, the image on the top of the page. Consequently, you change the image to something that you believe to be more attractive.
Lo and behold, your conversion rate rises to 1.5%!
This proves that your hunch was correct, right? Well, perhaps it does but maybe it doesn’t.
The problem is that there are a multitude of variables that could have caused this leap in conversion figures.
For example, the audience presented with the latest version of your squeeze page may be completely different to the audience who saw the original one. You can see this from your tracking stats but what you don’t know is how this new audience would have reacted to the old version of your squeeze page.
In other words, they might have pushed your conversion rate to 1.5% even if you had not changed the image. You will never know whether this is the case or not. You are not therefore comparing like with like. This is an ‘apples compared to oranges’ test whereas you should be testing an apple against an apple.
To produce valid statistics that really point you in the right direction, you need to show the two different squeeze page versions to the same audience. In this way, you can establish their preference which is the information that you really need.
In practical terms, this is unfortunately very difficult to do manually. You would for example have to set up two different URLs for your squeeze page versions and it will be very difficult to guarantee that the same audience cross-section visits both in equal numbers. This is why most professional online marketers move to using testing software as soon as they possibly can.
Fortunately, you can do the same because there is an excellent free software program that allows you to test as many different variations of your web pages as you want. The program in question is the free Google Website Optimizer (this link takes you to the training videos page).
Using this software, you create a webpage – for the sake of continuity, imagining that it is your squeeze page – before testing as many different variations of this page as you can think up against one another.
The simplest form of testing is to run what is known as an ‘A/B split test’. In this scenario, you change just one element of your webpage to see what difference this makes to your results.
For example, you would commonly start by changing the headline that you use on your squeeze page to see whether you can improve on your current conversion rates with a new headline.
The value of testing should not be underestimated because even the smallest changes can produce significant results. As an example, I was offering a report called ’10 Hidden Weight Loss Secrets’ using a squeeze page, using the Website Optimizer tool to test two different headlines for effectiveness.
The first headline was ‘Free Today – 10 Hidden Weight Loss Secrets’. The percentage of visitors who subscribed to my list from this squeeze page was 2.6%. However, the version that was identical apart from featuring ‘10 Hidden Weight Loss Secrets Free Today’ as the headline converted at 3.26%.
When you have created a web page, you can use the free Google tool to ‘A/B split test’ every element of your page in a constant effort to keep improving your returns.
You can also use it for multivariate testing where Google tests many different variable combinations at the same time to establish what works best.
The secret behind the effectiveness of this is that because all of the testing is conducted using the same URL, the audience for each different version of your offer is the same. Hence the only possible difference between the results for one version of your page and those for another is the reaction of that audience.
Establishing audience reaction in this way is of course the key to testing.
For most new online marketers who are just learning the business, tracking and testing are a long way down their list of priorities.
Misunderstanding or ignoring the power of tracking in testing is however a major mistake, one that often consigns marketing beginners to a never-ending struggle.
As previously mentioned, most established, successful marketers swear by the value of tracking and testing.
Is it any coincidence that these people are the most successful in the business whereas those who ignore tracking and testing make far less, if any money?
It is now recommended you proceed on to Part 14 – Creating Your Own Products Using PLR.