NONE: Re: ONLINE-ADS>> Web Tracking Solutions
Re: ONLINE-ADS>> Web Tracking Solutions
Brady Chatfield (bchatfie_at_pcshs.com)
Thu, 17 Oct 96 19:09:56 -0700
(snip some good comments on I/PRO, etc.)
> In terms of off-the-shelf, don't look for a pre-built cookie solution in the
> near future anyway since there is no standard way of writing cookies or even
> what kind of info is stored there. I think if you are smart enough to write
> information to cookies, you are probably smart enough to collect it &
> organize it when the user returns.
Well, you don't need to write detailed information to cookies; they're not
useful for tracking in this way. However, there are other applications of
cookies that are only now being explored, for example:
We simply send a unique character string so you can tag a browser as being
unique in relation to all others. The information the cookies have basically
says "I'm John Smith"... we can then reference anything on "John Smith" that
we have chosen to store from previous visits. Therefore, if the "John
Smith" cookie remains intact, I can always tell you with 100% certainty
when he visits. Standard logs cannot do this, no matter what anyone
tells you. It's one thing to know how many people visit your site, but
quite another to tell how many are repeat customers!
>
> My personal preference: plain old log files with plain old Interse. From
> our standard reports, we already get detailed reports on 404 errors,
> visitation to every page, path thru the site, referring URLs, user domain
> type, organization & geography, browser type & version, bandwidth usage by
> hours & days, and more. I understand that Interse will also track
> "registered" users, but we haven't tried this feature yet.
>
How can a standard log detail a path through site, and if that claim is
made, what is the accuracy of it? All you know about a hit in a standard
log is the source Internet address. Period. So, let's say your company
has oh, about 5,000 computers behind a firewall. Guess what: any one of
those browsers in there will give the same Internet (IP) address. What
sort of magical orb does Interse' have that can claim to detect these
frequent occurrences, which will really mess up a track? Cookies not only
accurately track a visitor through the site, they also can tie key events
(product purchases, authentications, etc.) to this visitor. This is what I would
consider a more realistic type of visitor tracking.
Geographical resolution is the biggest crock of bull I've seen,
especially in I-PRO ( I believe ). They say when you get 500 hits from
Microsoft (microsoft.com), that means that 500 hits are logged to
Redmond, Washington. Well, the local Phoenix Microsoft office's domain also ends
with microsoft.com. So, your tracking software will say these guys are in
Washington? Don't think so. If anyone tries to say their product can do this,
run, don't walk, and make sure to hide your checkbook.
...And 404 error recording doesn't belong in a "marketing analysis"
package. A 404 error is when you dial up a page and it isn't there. This
type of stuff gets recorded in the server error log by default, so this
is not a feature of a stats package as far as I can tell (or would pay
for).
Also, Interse's "register to track" scheme is a joke. People do it, but I don't
think it captures a real sample of the visitors coming through a site. With
cookies, visitors do nothing on their own to be tracked (well, except
for coming to the web site); the tracking is completely transparent to
them. As a marketer, here are the choices I present:
1) Take a sample of the population and interpolate what
data you need.
2) Your sample _is_ the population, so you have no need to
interpolate your data, hence no (significant) errors.
Choice #1 is having a customer register somewhere to then be able to
track them (interse'). Choice #2 is cookies.
> Nothing's perfect, but that's just about all you can glean from logs so far.
> And hey, who are we to complain when that is about 1000 times more than
> radio, print or television media know about their audience, without all the
> expensive surveys & Nielsen reports?
>
Yes, but if we *have* this great technology then let's use it to its fullest!
I don't believe that cookies in their present incarnation are the solution for
the future. They are a kludge solution in an attempt to maintain some sort of
state reference between a server and browser. But they are the best thing out
there. Cookies will be around for some time, so we should use them to their
maximum effectiveness. There are limits to their use, as in anything else;
however, I think people are going to catch on to using cookies more effectively,
and we will see this in more products.
Brady Chatfield
bchatfie_at_pcshs.com
