|
DO YOU
REALLY ANALYZE YOUR WEB TRAFFIC?
by: Cary Christian
John Q. CEO sat down at his computer just the other day and opened up the
website statistics package provided by his host to see how his website was
performing.
"Hmm," he smiled. "5,000 hits yesterday. Great!"
John quickly closed down the stats package and went back to a report he had
been reviewing.
Okay, I know most of you already know that John Q. really didn't learn
anything from his quick perusal of his statistics, even if he thought he
had.
But do you really know what you SHOULD be looking for? "There's GOLD in them
thar logs!" as the old prospectors would say.
FIRST, THE BASICS
Hits matter very little. One visitor opening one page on your site will
register a hit for every image, banner and JavaScript file on that page.
Essentially, if the page calls or uses a file of any type, it will register
as a hit. So one visitor visiting one page could generate 10, 20 or more
hits. Even if your pages are relatively free of images, banners and so
forth, you'll still likely get three to five hits just from your borders and
logo.
So if John Q.'s web pages have high graphic content and a lot of scripts
running, he might have gotten no more than 500 actual visitors.
Pageviews are a much better indicator of traffic. A pageview represents your
page being viewed as a whole entity. A UNIQUE pageview is a view of
your webpage by one person within a 24-hour period, so return visitors are
factored out.
But even this information is too much of a high-level indicator to be of
much use. Let's examine where the real gold lies.
THE REFERRER LOG
Where are your hits coming from? That's important. It is an indicator of the
quality of your advertising efforts. You'll be able to see which search
engines are sending you traffic and how much. You'll know when people are
clicking through from email marketing, from ads you've placed around the
Internet and how well your reciprocal links with other sites are working.
This is all information you need to know. It will help you refine your
marketing efforts and place more emphasis where it will do the most good.
What I really like, though, is the information the referrer log gives me
about the keywords and phrases people use to find our site. This is not just
good stuff, this is very exceptional stuff! Why?
Let me give you an example.
After tracking the keywords people use to find our site over the course of
the last year, I noticed that hundreds, and sometimes thousands, of people
were using almost the exact same phrase every single month. The phrase was
driving people to our site because of an article on that subject in our
"Business Articles" section. (Let me digress a moment: remember how I told
you articles and other content on your site could pull in new visitors? This
is an example).
I wondered why we were getting so much traffic from this phrase because of
just a single article. Logically, that must mean there is not enough
information online on that subject. (Trust me, it's a subject you wouldn't
think you'd get thousands of hits on!).
So I dutifully went to the search engines and started searching. Sure
enough, I found a real hole in the coverage on this topic. (Remember last
week's article on the importance of finding a niche? This situation is
starting to highlight a lot of different concepts isn't it)?
Needless to say, an ebook is in the works to give these searchers what
they're looking for. I believe it has great potential based on our traffic
for this subject.
I've also identified several other trends in what people look for when they
find us. Each trend might eventually become a new product. In a sense, our
visitors are telling us what they want and, possibly, what they will buy.
BOTTOM LINE
Spend some time really reviewing your log information. Almost all statistics
packages will provide you with a wide variety of information. A lot of it is
interesting but meaningless in terms of real profitability. But if you pan
for the gold in the logs, every once in a while you'll find a nugget that
makes it all worthwhile!
Copyright (c) 2003
|