Re: Measuring the Branding Impact of Online Ads
Arun Natarajan <arun_natarajan_at_vsnl.com> wrote:
>It seems that Measurability is a very important factor
>in this issue: while click-thrus are easy and objectively
>measurable, there's no easy--and universally acceptable--
>way for advertisers to measure the branding impact of
>online ads. As a result, advertisers tend to rely on
>click-thrus to evaluate effectiveness of all online
>advertising.
>
>Though most of us acknowledge that this situation is not
>v.desirable--for publishers, at least ;-)--the problem so
>far, has been the TINA ("There Is No Alternative") factor...
>
>I recently came across two firms proposing different
>approaches to solving this problem: one from research
>firm, Jupiter Media Metrix and another by AdIndex (a
>division of Dynamic Logic)
>
>* The JMM Approach
>
>The relevant JMM report is titled "Is the CPM Model Dead?:
>Making Pay-for-Performance a Sustainable Media Buying
>Model for Advertisers and Publishers". I'm extracting
>a couple of paras from the report beloew to highlight
>their proposed "hybrid CPM-CPA model" solution
This is no solution at all.
>* The AdIndex Approach
>
>According to the site (www.adindexnetwork.com), AdIndex
>is a tool that "aims to provide attitudinal metrics to
>online media campaigns and help advertisers measure the
>impact of ALL their impressions - not just those that
>were clicked".
>
>AdIndex uses online surveys to gather data from visitors
>to the participating web sites. Visitors are randomly
>intercepted to take a brief survey and, upon their
>consent are asked a series of questions designed to
>measure the communications impact of a campaign.
Better than nothing, but still inadequate.
My suggestion is that tightly controlled experimental
designs be used in selected markets in which exposure
from a range of media can be confined. The research
design should measure pre/post increment in awareness
and attitude. Media can then be meaningfully compared
in terms of cost of achieving goals. Media could be:
Local TV; Newspapers; Radio; Opt-in email;
Online Banners, Text Ads, Pop-ups and pop-unders.
An experimental design eliminates the effect of all
variables except the one(s) being measured.
Frankly I'm surprised that big ad spenders like P&G haven't
done this already!
Ivan Weltman
http://www.tudogs.com
Received on Wed Dec 19 2001 - 09:40:54 CST
