Online statistics tell the truth, the whole truth and nothing like the truth

Internet business owners rightly want to know how well their web site is performing. Thankfully virtually every web site hosting package these days comes with statistics. However, these statistics are often misleading.

Hits, for example, tells you nothing. You can still see some web site owners with little badges claiming a certain number of hits. But hits are about how many bits of data were collected by the user’s computer. This page alone, for instance, will lead to 52 hits. The number of hits tells me nothing – after all you could refresh this page three times leading to more than 150 hits, but if you read nothing that statistic is of no value to me.

Web site measurement companies replaced hits a while back with page views, the number of times each page as a whole was looked at. Even though this is more sophisticated than page hits, it is actually a rather useless piece of data. Once again, people can load your page, realise immediately it is not what they wanted and then depart within milliseconds. But you think someone read your information; the statistic is lying.

Unique page views is another change to web site statistics which was touted as an increase in the value of the page view information. The unique page view tells you how many different people logged on to your page. This overcomes the refresh problem, for instance, but also deals with one other situation. You could have 1,000 page views all from the same person, so you think you are 1,000 more times successful than you actually are.

The world of newspapers and magazines have used a piece of data called “page traffic” for several years. This is sophisticated, expensive research which shows exactly how many people read each story in a newspaper, how many people looked at each picture, each headline and so on. It tells them how far through a story readers got and how much time they spent with each item in the newspaper. This detailed information helps them make sure their newspaper is closely matching their readers’ needs.

If newspapers only had the total numbers of people who turned over the pages – the equivalent of online web statistics, their newspapers would fail to meet their target readers effectively. Is it any wonder that many web sites fail to target their readers well, if all they are using are basic statistics which really tell the owners nothing?

Now, the market research company, Nielsen/Netratings has come up with a new idea. They are looking at the total time people spend on each page as a measure of web site value. For instance, the total page view of web sites offering instant messaging is 0.4% of all page views in the UK. But these pages represent 13% of page view time, according to New Media Age magazine.

So there is clearly a difference between how many pages are viewed and how long people spend on them. The idea between the Nielsen/Netratings new measurement is that rich media such as video mean that the older page view statistic is less informative for web site owners.

But there is a problem with the time measurement. People can open a web page, realise it is not for them and go back to their search straight away. But with tabbed browsing, the web page they do not want could still be “open”. This means it will be recorded as being viewed, even when it isn’t. Equally, people may spend a long time on a web page because they are confused or can’t find their way around it. The time measure could therefore be a measurement of lack of success, rather than a measurement of how good a web site is.

So, how can you measure your web site’s effectiveness. The statistics can help you, but don’t rely on them. Even expensive web site statistics measurements have flaws and need considerable care to interpret them. They can easily lead you up a blind alley.

Instead, ask your users directly. Meet them, speak to them, survey them, use eye-tracking data to measure what they do. In other words, rather like printed newspapers who use page traffic data, you need to know exactly what your users do with your pages. Raw data is of little value to your online business; real information is much more important. Yes, I know that means more cost and more time. But what do want? Online success or not?

Like this article?

Share on Twitter
Share on Linkdin
Share on Facebook
Share via email

Other posts that might be of interest

Photo manipulation in progress
Internet Psychology

Do you airbrush the real you?

I am running the risk this week of being disciplined at work. If my boss reads this (and she often does) she will discover I have done something naughty. Recently, I posted a picture on

Read More »