When the client is king
Op-Ed: “If page impressions, hits, and hosts aren’t reliable in Internet research, why not rely on tried-and-true methods?” asks Peter Wälty.
news magazine Facts took up the position of online editor
had a story in the magazine with the title "Der gläserne
Surfer" was published. Every movement of an Internet user can be registered
his interests, preferences and habits. His consumer behavior
would be on a platter, ready to be taken advantage of by nasty
to be exploited by marketing departments. Tailor-made
Products should entice the surfer to buy. It turned out differently. It came
Spam - a napalm carpet of indiscriminately sent advertising e-mails whose
advertised products do not interest us for the very reason that
because we don't need them. Or who has a
Penis enlargement necessary?
There is a similar story to tell about measuring the reach of websites:
Total transparency was fantasized about. Webmasters measured what
the log file revealed. And that wasn't very much. Soon enough
It turns out that hits (retrieved files) and
Page impressions (pages viewed) are completely unsuitable measures
because they are always dependent on architecture,
Layout and offer of a page and have with range about so
much as page numbering has to do with reader numbers.
Then, in analogy to print, an attempt was made to increase readership or visitor numbers.
determine. From now on, hosts - computer addresses that can be found on the site
were registered. A nonsense: hundreds of Tamedia employees
appear to the outside world as exactly one single host - namely as a
Tamedia proxy server.
Since 2000, Nielsen Netratings has used a panel to register
visited sites - but unfortunately only for private users. Then the
Wemf brings order to chaos with NetAudit and establishes a standard
the Unique Client (UC). However, the UC is by no means
a single user, but in the best case a single PC. If
I use the same website in the store, at home and with my girlfriend
visit, I am three UCs. When a family of five on one PC
visits the same website, all five are exactly one UC. And if with
the NetAudit cookie is deleted after each shutdown, I am - if the
value is not corrected by a nebulous algorithm - on the
a new UC the next day. So distortions are literally programmed: Sweden's largest daily newspaper on the web
measured in this way, shows more users than Internet users in the
country exist.
With the New Internet Research Switzerland - a combination of
measurement and questioning - became something of a silver lining on the
horizon visible. It would be unfortunate if this project were to pass the
financing would fail. Regrettable, but not a disaster, because
In my opinion, there is a valid alternative. Print titles are
measured with telephone interviews for decades. Why not actually
Websites too?
Peter Wälty is head of 20minuten.ch
