Re: online targeting

From: Clint Ballard <clint_at_accelerationsw.com>
Date: Thu, 4 Feb 1999 08:08:51 -0600 (CST)

TLEE_at_WEBCMO.COM WROTE:
> The real problem of this targeting and not targeting
> comparison is not about numbers. It is about how we use
> numbers to help our decisions.
>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

If campaign A costs $143 per sale and campaign B costs $500
per sale, what else really matters? How is this NOT about
numbers? In online marketing, it is best if you base your
decisions primarily on the numbers. When things are changing
so fast, you can never be sure about much of anything other
than how much did it cost to get the sale. If you are sure
that the initial sale will be followed by subsequent sales
of significantly different levels from the two groups, then
your lifetime cost for the sales need to be adjusted
accordingly.


TLEE_at_WEBCMO.COM WROTE:
> Usually when we talk about targeting, we mean to expose the
> banners to the people who are more interested in the product
> (or service) in general. However, if we follow this case
> from Iconocast, we need to find those people who have LESS
> INTERESTs in the product (or service) than the general
> population but are MORE LIKELY to buy the product.
>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


I thought we are trying to get conversions at the lowest
cost, not just showing banners to people who are interested
in the product. The Iconocast result does not imply that we
want to find people who are less interested, just the
opposite, since the conversion rate was 30 times! It just
happened that the CTR was lower than the Untargeted group's
CTR, this doesn't mean Targeted was less interested. It
could mean that the population of people who converted 30
times more are VERY busy and as a group have a significantly
lower CTR, regardless of their interest level.

It could also mean that impressions were counted in a very
DIFFERENT way! I am not familiar with the specifics of the
campaigns to know if the two campaigns used identical
versions of the same ad serving software with the same
configurations having placements in identical locations on
the screen with the same frequency and identical browser
targeting options shown at the same time of day. Unless ALL
of these parameters are the same, you cannot make a
meaningful comparison of CTR, since nobody actually knows
how many times a specific banner has been shown. Last year,
I ran a parallel campaigns using two different ad servers on
three different websites using the same banners at the same
time and got a performance difference of 82%, 140% and 3000%
(yes that is not a typo) respectively. A difference of a
factor of two from the Iconocast result could easily be
explained by this type of variance.


TLEE_at_WEBCMO.COM WROTE:
> Are we able to do that?
>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


We don't need to do that. If we measured the cost per sale
after taking into account the different CPM's (regardless of
how the impressions are counted), CTR's and conversion rate,
then that is what matters. Why do we really care about the
specific relationship between the CTR's (or CPM's for that
matter)?


TLEE_at_WEBCMO.COM WROTE:
> Numbers don't have any meaning unless we give them the
> meaning. So when we get some numbers, we need to be cautious
> to interpret them. Especially your interpretation may be
> used by many others in their marketing decisions.
>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I couldn't agree more! The numbers 143 and 500 by
themselves are meaningless, but when we say that those are
the two different costs per sale, then it is safe to make
marketing decisions based on those numbers. Marketing
decisions should not be made on the precursors to the cost
per sale, or whatever the ultimate goal is. Also, after the
initial test, you need to continuously monitor the
performance to make sure that some critical factor hasn't
changed.

Clint Ballard
C.E.O., Acceleration Software, clint_at_accelerationsw.com
"http://www.clicksales.com/index.html?oa", pay-per-click advertising with a
100%
satisfaction guarantee


========================================================================
------------------------------------------------------------------------
       This week's Online Advertising Discussion List sponsor:
                                Conducent

   Don't Buy Hope! Get a 94% User Response Advertising with Conducent!
     Place Dynamic Ads in Downloadable Software. Invest in Results!
 http://www.conducent.com/advertisers.shtm or eisenstadta@conducent.com

------------------------------------------------------------------------
========================================================================
Online Advertising Discussion List To Unsubscribe send UNSUBSCRIBE
http://www.o-a.com/ to online-ads-request@o-a.com



Received on Thu Feb 04 1999 - 10:54:04 CST