Do you have great numbers of well qualified traffic coming to your website but disproportionately low numbers of sales? How about paid registrations? We have worked with a number of customers who have the same issue. Upon reviewing their websites there doesn't appear to be anything wrong at first glance. The sites look professional, have nice graphics, seem to load well and work the way they're supposed to. The copy is well thought out, the content is well targeted and the sites have good unique selling propositions. On top of all that the offers are backed by 100% guarantees along with testimonials from satisfied customers smattered throughout the pages. So what on earth is wrong? Why do we see websites with ridiculously low sales conversion rates at often less than 0.5%?

Getting higher than 0.5% sales conversion rate

A well thought out website might hit 0.5% or higher anyway without any extra work. More often than not however problems exist within a process which forces the majority of visitors to work harder than they want to. It really is that simple. If you want to improve your conversion rates you need to find out which processes or steps are causing the problems before people can actually convert i.e. buy or register. By measuring how people use your website in particular the common scenarios and paths they follow then you can start to figure out what your problems are.

Introducing the web scenario

Most decent web measurement systems these days allow you to set-up a scenario that is likely to be frequently used by your website visitors. Take a typical shopping cart scenario. Your visitor might arrive at your campaign landing page then add one of your products to his shopping cart. After this the ideal scenario for you is for the visitor to continue by filling in his credit card details and confirming his purchase. This means you might have a scenario which follows this path through your website.

Page 1) product_landing_page.php

Page 2) add_to_cart.php

Page 3) review_order.php

Page 4) add_details.php

Page 5) purchase_success.php

What typically happens is that people arrive in large numbers at page one but only a small percentage end up at page 5. This is a process and it's what needs to be measured to find out where people are leaving from the website. If you can determine where and how people leave your shopping cart process (also called shopping cart abandonment) then you can attempt to do something about it.

Shopping cart abandonment

Let's assume that you have a similar process to the one described above and you have a measurement system that allows you to see how many people visit each page. In two of our recent cases we have seen statistics which resembled these figures.

Step 1) product_landing_page.php – 10,000 visits.

Step 2) add_to_cart.php – 8000 visits (20% abandon)

Step 3) review_order.php – 240 visits (97% abandon)

Step 4) add_details.php – 120 visits (50% abandon)

Step 5) purchase_success.php – 40 visits (66% abandon)

Let's examine those figures in a little bit more detail. Firstly 10,000 visitors arrive at the landing page. About 20% of those visitors leave from the landing page without adding anything to the shopping cart. From the remaining 8000 who added the product to the cart only 240 have gone one step further and reviewed the order (3% of the remaining 8000). After that a further 50% left the process when you asked them to give you their credit card details and a further 66% have bailed out at the last point. This means that from a potential 10,000 purchases you have only 40 buyers. A pitiful 0.4% conversion rate.

Examine the biggest bail out rate

Suddenly you can see why even though large numbers of traffic have arrived at your landing page only a small percentage buy and the biggest problem is because of one part of the process between step 2 and step 3. If you can reduce the number of people leaving at this point to lets say 50% (very doable by the way) then your figures would now look like this:

Step 1) product_landing_page.php – 10,000 visits.

Step 2) add_to_cart.php – 8000 visits (20% abandon)

Step 3) review_order.php – 4000 visits (50% abandon)

Step 4) add_details.php – 2000 visits (50% abandon)

Step 5) purchase_success.php – 660 visits (66% abandon)

That would mean a fantastic increase in the numbers of buyers. Instead of a 0.4% conversion rate you now have a 6.6% conversion rate. You basically have the same advertising spend to drive the same numbers of traffic but you have a much improved return on your advertising spend. This is one reason why measurement is so important because without measuring the scenario you would never know the problem existed.

So how do you improve the scenario?

That's the million dollar question and in some case we've worked on it really can be worth a million dollars or more. The follow up article to this one is going to show one such improvement in detail however it's impossible to say why people might be leaving your site in similar numbers without a detailed study of the problem pages. We've found it's usually because the website is making things un-necessarily difficult for the website visitor. They might have scary looking forms to fill in, or moving from the shopping cart to purchase page might not be obvious. The point is that web analytics systems can point out these problems so that you can attempt to figure out what is wrong.

In summary web analytics does 2 things well

Using web analytics tools allow you to establish facts and begin developing scientific tests. Firstly you should establish the facts so you may find out where your problems are as we illustrated in this article. Then once you have determined where your problem areas are you can begin working on improvements. If your improvements make it easier for your visitor to move to the next step of your purchase scenario (in this case) then your next test should try to improve the next step and so on until you have the abandonment rate as low as possible.