Re: ClickTracks

From: Jim Novo <jimnovo_at_sprintmail.com>
Date: Mon 10 Feb 2003 19:30:07 -0600

Matt Mickiewicz typed:

> It's definitely a very unique log file analysis program. It's more
>intuitive than looking at a pie chart, graphs and a table with a whole
bunch of
> numbers.

Matt, I respect your work at SitePoint and you are always on track with
your advice. On this one, I think there needs to be some clarification.
There are different tools for different needs, and I get concerned when
people make broad statements about visitor analysis software or services
without qualifying what the analysis user is trying to accomplish.

If what the user wants to do is look at generic visitor information and a
particular interface makes it easier to get to the info they want to see,
that's great. There are tons of tag-based programs out there that do that
in different ways. But usually, the information reported on with these
services is what *they* (the analytics company) think you want to see - all
the generic reports. I have yet to see a tag-based program that allows me
to get the information *I want to see*. I want to choose my reports, not
have the choice made for me.

Is there a tag-based program that allows me to see the fall-off at each
stage of the shopping cart by Search engine by Entry Page? How about
newsletter sign-ups by Search Phrase? Downloads by Pay-Per-Click source
code? Items added to the cart by path? Understanding this stuff is how
you make more money. I don't have a single client that cares about "page
views". They care about figuring out exactly how to make more money. They
never even look at the "standard" reports, because they don't lead to
making more money.

> ClickTracks does the same thing as WebTrends - it just displays the
> results in a much more meaningful way, in my opinion.

The above is an example of exactly what I mean. If all you want to look at
is generic visitor information like the standard WebTrends reports, and you
like the ClickTracks display better, go for it. Personally, I never look
at any of that stuff and I set up WebTrends to not even bother running
those reports. For any given client, I know very specifically what data is
the most important to making that client more profitable, and I can't get
at it any simpler or faster than I can with WebTrends.

And I'm not talking about a $50,000 dollar piece of software. I'm talking
about the basic Log Analyzer product you can *own* for $500, which also can
pull in non-web data through an ODBC connection. So, if you have external
info that matches anything in the log data (like log-in name or
"authenticated user"), you can pull that info right into the visitor
reports. Can that be done with a tag-based solution? Personally, I would
much rather know the specific visit patterns of my highest spending
customers than how many random visitors entered the site through a certain
page. How do you make more money knowing that?

By the way, the new Webtrends Reporting Center 6.0 (next jump up from
Analysis series) combines both standard server log analysis and tag-based
analysis into the same interface and reports. This provides the advantages
of both approaches and removes the disadvantages of either - for example,
try tracking downloads or 404's with a tag-based program. Can't.

Hurol, if you're still with this conversation, WRC 6.0 has visitor Recency
(days since last visit) and Latency (average days between visits) built
right into the reporting. I know you talked about some of the these ideas
in your book published last year, as I did in mine back in 2000.

Using these metrics, users can now get into things like average days since
last visit by ad source, allowing them to rank advertising ROI by the
long-term value of the customers generated instead of just by initial
conversion to sale. It used to take days worth of intense SQL queries to
get this info back in 1998. Now anyone can get this info in minutes,
without the need to be a programmer to make it happen.

Progress!

Jim

Jim Novo | Author: Turning Customer Data into Profits
http://www.jimnovo.com
Co-Author: The Guide to Web Analytics
http://www.hiqhq.com/webmetrics.asp





Received on Mon Feb 10 2003 - 19:30:07 CST