The types of conversions remain the same across the life cycle of your web site, but the actual con-
versions may not. And that’s the reason you need to conduct regular conversion analysis. In your
conversion analysis, you’re looking for a number of things including an increase or decrease in the
number or pattern of conversions. Increases and decreases in these trends will help you to know
when something is working or when something needs changing.
When you see a changing trend in your conversion rate, you know it’s time to begin trying some-
thing different. It may be that your products need to be changed, it may be that your newsletter is
no longer appealing, or it could simply be that your competition is funneling your audience away
One way to see these trends in your web analytics is to look at your sales cycle. Ideally, you should
know your sales cycle even before you put your web site online. The sales cycle is the steps that a
user takes to complete a sale. It’s usually divided according to the natural progress that visitors make
from entering your site to the completed sale.
This natural progression should be trackable with your web metrics program. One report that will
be extremely helpful will be a funnel report or sales funnel report. This type of report allows you
to set up a sales funnel on your web site and then track the effectiveness. It also shows where your
site visitors might fall off in the process of making a purchase. You can then use that information to
change and improve elements of your SEO and your web site that might be causing that fall-off.
Conversions are the ultimate goal in all your search marketing efforts. Yes, you want to drive cus-
tomers to your site, but more than that, you want those customers to do something. That something
is your conversion and if you analyze the conversion process, you’ll be able to track patterns of what
works and what doesn’t.
Analyzing Server Logs
Web server logs are the records that servers keep of who accesses files (in this case, web sites) that
are stored on the server. A server log is a long, ugly, and complicated document in which lots of
very usable information is buried. But if you don’t know what you’re doing, it could take a very
long time to figure out what’s usable and what’s not in that server log.
That’s not a huge problem, though. There are all kinds of server log analyzers that you can use to
help you decipher the log and find the information that you’re looking for. One of the most well-
known is AWStats (
). An example of the Robots/Spider
Visitors report is shown in Figure 21-1.
One thing you’ll notice about AWStats is that it looks a lot like the web analytics software that
you’ve looked at. That’s because it is essentially the same thing. Web analytics software does the
same thing that a server log analyzer does. The difference is that server log analyzers sometimes
provide more information than basic web analytics applications.
21 1 9:59 277