Link Building Blog, News and Articles - Textlinkbrokers.com/blog

Blog Homepage : Category - Marketing Tools

Most websites have some sort of tracking whether it is a real time analytics package, or a log analysis based webserver logs, analytics should form the foundation of any online marketing campaign.  If you are using analytics you MUST read this.

In order to understand what analytics to use you need to know a little about your system.  But that’s ok if you don’t know for sure.  Usually your hosting provider can help you figure that out.

Basically there are 2 types – one is the more traditional log based system, that uses your webserver logs and analyzes them to produce reports, while the other is an on-demand system which utilizes code usually placed on each page of the website to track visitors and output reports.

Both systems have positive and negative aspects which could impact your decision of which to use.  And I can’t stress it enough – you MUST use analytics of some kind if you have a website.  Otherwise you will never know where your visitors are coming from or what they are doing once they hit your website.

Log based systems are probably the most well known.  They come in different flavors – from a free version, such as AWStats to a paid version, such as WebTrends or Urchin.  While different systems rely on log files to produce the reports, there are different qualities of reports and different details that go into those reports.

Generally, I prefer log based analysis mostly because with logs you get all traffic, and not just those that can execute JavaScripts.  This is because usually with the on demand analytics, the site visitor must execute a script in order for the system to record the visit (I’ll get more into this later) whereas with log files, all traffic regardless of their script executing abilities get tracked.  This means the search engine crawlers will be tracked more effectively than with some realtime analytics.

One drawback, however, of the log based systems is that they tend to be resource intensive, especially if you need to run the reports locally, and they have to be manually set to run on a regular basis. 

Another drawback, with some packages, is that they lack the detail you may need to make an informed decision.

For example, a basic install of AWStats (as hosted by many hosting companies) really only give you such simple details as the number of hits and/or bytes transferred over a date range, and what pages were requested. 

This despite the fact that most log files can be configured to log much more detail.

Ideally, your log files should be configured to capture as much detail about the visitors as possible.  Even if you don’t use it now, you may want to use it later.  The problem you may run into is that the host may be unwilling or unable to log that much data, and they may charge you for the extra space such logging needs.

For example, if you log everything, and want to store your logs for 6 months, there could be a substantial charge as these logs could use many megabytes of disk space even if they are compressed.  Of course the other option would be to download them and store them locally, then you wouldn’t have to worry about paying for extra space, but you’d have to remember to download them as many hosts will delete last months logs a few days into the next month.

A program like WebTrends has great analysis abilities.  WebTrends is one of my personal favorites because once you get to know what you can do, it’s quite easy to set up multiple custom reports.  However, if the site gets a lot of traffic, generating reports can take a lot of time.

On demand analytics

These types of analytics are becoming much more popular mostly because they can provide you with up-to-the-minute data on who is visiting your site, how they got there and what they did once they came to the site, regardless of how much or how little traffic the site gets.

And rather than configuring multiple reports and re-analyzing the data over and over (as is common with log based analysis) most of your reports are available within a few mouse clicks.

Even custom analysis can be performed in a fraction of the time it would take to define a log based report, and run it to analyze past performance.

With most real time analytics, it’s merely a matter of selecting the report you want to see, and picking a date range, and viola, a few seconds later there is the report.

Most of these types of programs also allow for custom reporting.  For example, if you wanted to know how many of your sites visitors came during a specific time period, and from a certain geographic region, you can easily find out.  Then if you wanted to find out what paths they took, you can define that, and finally if you wanted to append your path report with a conversion report to see what they bought and even how much it cost, you could do that too.

As mentioned above, there is one drawback and that is that many of these programs are script based which means that the visitor has to execute a JavaScript in order for their visits to be recorded by the analytics package.

If the visitor won’t accept scripts (like a search engine crawler won’t) then you won’t get to see their traffic.

Of course, some packages try and get around this by attaching the tracking code to an image, for example, and recording the accesses to that image.  In this case, the image request that comes from a search engine crawler would be recorded just as a request from a site visitor would.

Or the package may require a visitor to accept cookies, but again a search crawler won’t accept cookies so the crawler traffic won’t get recorded.

In a nutshell those are the main differences between the 2 most common forms of visitor analysis.  Now its time to examine what you need to get out of your reports.

What to Look For

First is to define what you want to see.  And to do this you need to know what the purpose of your website is.  For most site owners it is sales made, but there are other things you may want to track such as email addresses stored, catalog requests, PDF downloads and more.

Think of every way you could capture someones information – this is a potential future customer and one you may want to follow up with later.  Therefore if there is a form to fill out, this too would be information you’d want to track.

Once you have decided the multiple conversions you want to track (in this industry those things I listed above are considered conversions) it’s time to do some homework and see which analytics package(s) will return the reports you are looking for.

With most real time analytics, you can insert custom code which allows you to record these conversions and even output them to a nice simple report.  With log analysis it may be a little more difficult, but not impossible. 

One thing I’ve done in the past is set up custom “thank you” pages for a form submittal, or PDF download or other conversion.  This way, the analytics package can track the number of times the special thank you pages were requested and using log analysis I can see how many times the thank you pages were loaded, indicating a conversion.

[b]Some things to watch for{/b]

One common thing I hear is “my site gets 10,000 hits a day!” Wow that sounds like a lot, but when you think about it, that many hits may not be that great.

This is because hits in all analytics packages refer to the number of files requested in a day.  This includes not only the page requested but also every image, stylesheet, javascript, flash movie, pdf, word doc or spreadsheet that is available on your website.

To put it in other terms, lets say it takes 20 images, 3 javascripts and one stylesheet to render your page.  When you factor in the actual page, that adds up to 25 hits – just to display one page!  That means that in reality, you’ve only have 400 page views.

Let’s take this a step further.  Let’s assume that your average visitor views 4 pages per session.  That means you’ve only had 100 visitors in one day.

If you have a 2% conversion rate (that is, 2% of all your visitors buy something) you have only had 2 sales today.  If you average $10 per sale, that means you’ve only made $20 today.  Doesn’t sound as impressive huh?

And that’s the thing with analytics – you really need to understand what you want your site to do, and find an analysis package that will support you.  Otherwise it’s just numbers with no real meaning.

Analytics are also a great way to discover new keywords you can position your site for.  Chances are, when you review your keyword referrals, 90% of all traffic will come from phrases which you aren’t optimizing for.

But, if you are getting traffic from Google for “blue right handed widgets” there’s a good chance that with a little optimization for that phrase, you can also capture top spots on Yahoo! and MSN for that phrase.

And if you are into PPC, then analytics also could provide you with some undiscovered keyword markets which you can buy cheap and reap the rewards.  Chances are “blue right handed widgets” has few bids, so you can own the top paid spot for a few pennies per click.

When you combine this with the fact that you likely already rank organically for the same phrase, you now own the top 2 spots most visible to someone who may be in the market for a “blue right handed widget.”

As you may see, analytics can be a very powerful tool when used properly, and I haven’t even got into the problems they can solve.  That will be the topic of my next article.