Web Measurement: What You Don?t Know Would Make A Great Book


"What's in it for me?" you ask. "Why should I measure how people use my website? How does it help and what does it all mean?" The purpose of this article is to try to give you some insight into effective web measurement and to talk about the most important page of any website, the landing or home page.

Why measure at all?

Fred Flintstone lived in the Stone Age but we live in the Information Age. We deal with a constant flow of information from TV, websites, email, RSS feeds, mobile phones, PDAs, radio, newspapers, flyers, billboards, and magazine covers. Even the sides of buses hit us with information about companies, products and services. So why on earth, in the midst of this information overload, would you want to measure how people use your website, another source of data to barrage you with even more information? The answer is quite simple and is summed up best by the 18th century writer Sydney Smith. "What you don't know would make a great book."

Consider this.

Your business is selling $50,000 worth of product a week (5000 units a month) through your website. You are delighted with these results, as many would be, and you only measure them because you figure you're doing something right. However your competition, always watching and waiting for their chance, come along suddenly and steal a lot of your market share before you know what's happening. How? They were consistently testing how they could improve their conversion rate online and after they had maximized their conversion rate, they went out and aggressively targeted your potential customers. The bounders! However since the conversion rate on their website is much higher than yours, they eat your market like a hungry lion.

Let's put it another way.

You are successfully selling 5000 products per month through your website but your conversion rate is only 0.18%. According to research carried out by shop.org, the average sales conversion rate is 1.8%. That means that you could be selling 10 times as many products (50, 000)! Imagine what that could mean to your bottom line. If you don't know what your conversion rate is, then you don't know how to improve it or even that it needs improvement.

Measuring conversion is not complicated.

Measuring sales or prospect conversion is very easy. Over a given time period, you simply need to know how many people buy or register an interest in your product or services as a percentage of how many visitors turn up. However, there is more to effective measurement than simply measuring this kind of conversion.

What a good measurement tool should give you.

The ability to improve your conversion rate depends, at the very least, on 2 basic things. In essence, this is what you 'have' to measure to begin a conversion improvement program.

    * Firstly, you need to be able to accurately measure the number of visitors arriving at your website.

    * Secondly, you need to be able to see how they use the website by looking at the paths they have taken and how long they have spent browsing your pages.

Don't just sit there going hmmm?.

You look at the paths that regularly 'don't' lead to a conversion and try to improve them. Don't simply sit there looking at your path tracking tool wondering to yourself why people don't convert, but look at your website and physically use the path that your visitor has exited. This is where careful analysis is required and where comparisons should be made with paths that 'do' convert people. In many cases, variables that are present in the higher conversion paths are not present in the lower conversion paths.

It's that simple. If you regularly compare the best paths and the worst paths whilst measuring your changes consistently, there should be a steady improvement in conversion. You undoubtedly will make mistakes, but that is why you should carefully measure any changes you make, and why you should measure one change at a time. If you change more than one variable, then you won't know which change made the difference and you won't learn anything valuable.

Of course this takes a lot of time and effort on the web marketer's part, but I never claimed it was going to be easy. In comparison to say direct mail marketing or TV advertising, it is still much less expensive when you do make a mistake.

The landing page

The landing page deserves special attention. When people do a search on Google, for instance, they have something in mind when they get to your landing (home/index) page, and if you're not it, they have gotten to you by mistake. There is nothing you can do about this at all. It's a simple fact of life that people using keywords like "improving conversion" could be talking about a web site marketing campaign or catalytic converters for their car.

The landing page however does require special attention from you as a web marketer because you want to reduce the number of 1 page exits from this page as best you can. This means your focus should be purely on the visitors who arrive. How well you service their needs when they find you is critical to your level of conversion. Again, measuring the visitors who arrive and the ones who leave immediately (the bounce rate as it's sometimes called) is a good measure of how good your home page is at getting its message across. The ones who read for a few seconds and leave aren't your target market so don't worry about them. On the other hand, the ones who read for a little longer and leave might be slow readers, or might be your target market so concentrate on getting that number down. Your conversion rate for your landing page should rarely be measured as registrations or sales. It's more likely reading time (for those websites that make the proposition on the landing page) or click-through to another section of the site.

Here's an example...

Using our measurement system, we recently made a study of how people used our website. We found that the landing page was converting 68% of the readers. The objective of the landing page is simple: get the reader to move to another page. The landing page headline is "Are you driving qualified traffic to your website but not getting enough customers or prospects?" This headline, the fact that we go on to describe the target visitors dilemma in the first paragraph, and the fact that there are links to articles which educate the reader (more headlines, to pique the curious among you) mean that we get a good percentage of readers who arrive and continue further into the site. We're always working on the other 32%, but by analyzing the bounces, we found that 50% of them were possibly irrelevant traffic. We have an article starring Winston Churchill that describes how colorful language can grip a reader, and many visitors were arriving at our pages looking for a history of the great man. And as we mentioned above, we also found that some readers were looking for catalytic converters (the keyword conversion brought them to our website). So overall, it meant that only 16% of our target audience left without doing anything. Maybe the phone rang. We can't measure that!

Our tests on the landing page have been numerous, but now we're frightened to change the headline. Seriously! Because simply by changing the landing page headline, we improved clickthrough by 36%. That's nearly double what we were getting over the same time period six months ago. So if you think you can write a better headline for us than the one that currently grabs the attention of 68% of our readers, email me and I'll test if it is better than the one we're using!

Another thing we tested was urgency. We had a section on our landing page that said you could get a free e-book by subscribing before a given date. The date was cunningly set to change every day to the same day's date. It worked. We got high numbers of subscribers in a short space of time and hit a 35% conversion rate, which we considered incredible. Over 1 in 3 people subscribed to get the book. Why did we stop? We listened to our readers who were getting annoyed by irrelevant information on the landing page. New subscribers didn't mind seeing the message, but the returning visitors, the ones you should really pay attention to, complained about the same message with an updated date. It proves one thing though. If you have a special offer in mind, urgency works.

Incidentally, the fact that all of the above was tested on the landing page doesn't mean you should forget about the rest of your website. For instance, one of our recent articles is very well visited and got terrific feedback from critics and other web publications. But as an entry page, the URL also has a very high 79% bounce rate. We have analyzed it and have drawn a tentative conclusion. We think it's because we haven't given readers anything to do when they finish reading. They get to the bottom of the page and that's it. The end. Article over. And they leave. So now we're going to add a new section at the bottom of our articles which encourages subscription or clickthrough. Again, by analyzing and changing things, we hope to improve. If it doesn't make any difference, or in fact makes the rate worse, we have lost very little, we simply put the page back to the way it was. Testing is about trial and error.

In Summary

I will never be too clever to ever stop measuring how people use our website. I don't know what will work with our visitors the first time around. I couldn't have said that one headline would work better than another until I tested it. I couldn't have said that using great copy that instills a sense of urgency in the reader would work better than not instilling urgency in the reader until I tested it. I couldn't have told you whether adding article links to the first landing page would improve click through until I measured it. I couldn't have said whether one graphic would work better than another until I measured it. I couldn't have told you that all these small changes altogether would improve our subscription rate to over 15% every month, until I measured it. In other words, by measuring how people use your website, you can continuously improve it and therefore improve the conversion rate, which eventually has a positive impact on your bank balance.

In other words what you don't know about your visitors movement through your website would make a great book.

About The Author

Steve Jackson is Editor of the Conversion Chronicles, (http://www.conversionchronicles.com) and CEO of Aboavista a Finnish company which improves web prospect and sales conversion rates.

[email protected]