0

The Virtual Shopfront

Written By: Mark Steadman

The Virtual Shopfront

If you’re stood behind the counter of your shop, you know roughly how many people have visited on a particular day, how long they stay for, what they’re looking to find, and how many people ask for assistance. This feature comes as standard in all shops, and is available at no extra cost.

Your website however, runs by a different set of rules. Because your website isn’t a real place, you can’t physically seethe visitors.Instead you need a tool which can do that for you. There are lots of tools available, and they all have different names, but they can all be classified under the common banner: “web statistics”.

Counting Down the Hits

The most primitive form of web statistics is the hit counter. This is akin to the caveman’s club: crude, archaic and not particularly versatile.They can usually be foundat the bottom of therather less upmarket websites, andcontain a string ofnumbers that are supposed to tell the webmaster how many people have visited the site. Unfortunately, they don’t do anything like that.

Hit counters are almost completely useless for various reasons:

1. All they will tell you is how many times a given page has been accessed. The word access refers to all sorts of things, and a “hit” will potentially be registered each time one of the following happens:

1.a Someone visits Page X
1.b Someone follows a link to Page Y, then clicks back to Page X
1.c A search engine robot (like Google’s GoogleBot) reads Page X
1.d A directory editor (a human person) visits Page X to make sure it is suitable for their directory
1.e The web designer checks that the hit counter is working
1.e A link checker programme visits Page X to make sure all the links on that page are OK The problem is that the hit counter doesn’t discriminate between any of these behaviours. It just rolls up to another number. It also doesn’t make note of how many people are revisiting the site, how many hits are original, and how many hits are generated by actual people, not software.

2. Why, if I walked into a book shop, would I want to see a sign on the door saying “1,508 people have visited this shop since 1998″? What possible relevance does that have to me? All it does is give the shop owner a false glow of happiness, because he thinks that 1,508 different people have visited the shop, which is almost never the case.

3. They hardly ever look part of the furniture. Most webmasters tend to opt for the yellow on black LED display, regardless of the site’s colour scheme. Because they’re copy-and-pasted from a website (ie: they’re not bespoke) they won’t properly fit in with the site’s colour scheme.

4. They’re intended solely for the website owner’s eyes, yet they’re visible slap bang on the home page.

5. They’re showy – they try to say to the visitor “hey look at us, we’re popular”. If you’re popular, you don’t need to tell people how popular you are

I’m not daft; I know why people opt for these one-size-fits-all counters: they’re free and they’re easy. It’s just a matter of a quick copy-and-paste job to get a hit counter working on a web page, and it doesn’t cost a penny. That however, is their only virtue. The results are misleading and they look shoddy. So what’s the alternative?

Here’s the Science Bit … Concentrate!

OK, I won’t lie to you. The more advanced web statistics tools cost money – but not as much as you might think. I’ll get on to the prices later, but let’s look at what we can do with a little bit of cleverness.

Advanced web statistics tools can monitor and breakdown a whole host of data. Like a hit counter they can tell you how many hits you’ve had, but the better ones can also tell you how many of those were people, and how many were machines. The commonly found pieces of information collected by these tools are

1. Visitor’s country

2. Visitor’s web browser name and version

3. Visitor’s operating system name and version (ie: Windows 95/98/2000/XP, Linux, Mac OS etc)

4. What page the visitor was viewing previously before they clicked on your link

5. The time the visitor logged on to your website

MSO Media offers fairly advanced statistics which can show you even more:

1. What search engine the visitor used to find you

2. What keyword(s) they typed in to find you

3. Which search engines and keywords bring you the most visitors

4. Peak times, days, months

5. How often a page crashes on average

6. An estimate of how many people have added your site to their Favourites list

7. The first page a visitor sees when they come to the site (“entry page”)

8. The last page on your site they visited before moving on to another site (“exit page”)

So now we can find out how people have found your site, how long they stay for, what page they see when they first come in, on what page they decided to leave, when the rush hour traffic arrives, how many times the site goes wrong and how many people would want to visit your site again.

How Do They Do That?

Advanced web statistics tools, like those offered by MSO Media are “server side” tools, which mean they sit on the computer that is hosting your website. If your web server is worth its salt it should be collecting all sorts of data about your visitors, and putting this information in a bunch of log files. Web statistics tools read – or “parse” – that information and display it to you in all sorts of meaningful ways, often using graphs and charts to break it all down. Don’t worry, although it sounds sneaky it’s all above board, and doesn’t count as Spyware or anything like that. Logging is totally common practise with web servers and should never cost anything extra as it’s built into the software. The bit that costs money is the breakdown of the results. You try wading through a web server’s log file and you’ll soon see why I’m writing this!

Why Use Web Statistics?

Web statistics are essential to any website that advertises itself, because you can find out which advertising methods work the best by checking the referrals, ie the pages that link to your site. You can break down the results and find out

1. How many visitors you’ve had from Google AdWords campaigns (as the referral address for Google ads is always the same, no matter what page your ad is displayed on)

2. How many people followed a link in a signature on a forum (a great way of getting visitors)

3. How many visitors you get from your reciprocal link campaign

4. How many people are typing in your address or selecting it from their Favourites list

Also, when you know what pages people exit on, you can look at them to see why they’re driving people away. If it’s a page full of external links, add a “target” attribute to each link and set the value to “_blank“, so for example, a link to www.msomedia.com might look like this:

target=”_blank”>Web design by MSO Media

What the target=”_blank” bit does is forces the web browser to open this link in a new window. This is common practise with links pages, because it doesn’t drive visitors away from your site, as your links page is still open in the original window. It also means that your web statistics software won’t register that as an exit page.

The Moral of the Story

Get yourself some statistics – you could be flying blind without them. Don’t get tempted by hit counters as they’ll show you nothing useful. And like I said earlier, you don’t have to break the bank to get good statistics: for example, we only charge 30 per year for our stats, which gives you all of the information I’ve detailed above, and more!

About the Author

I’m Mark Steadman, director of MSO Media. My weekly eNewsletter “Getting Results” helps people get the most out of the Web, with practical advice, tips and tricks. Sign up at http://www.msomedia.com/nletter_sub.aspx.

Previous post:

Next post: