[Previous] [Next] - [Index] [Thread Index] - [Previous in Thread] [Next in Thread]

Subject: UKNM: Standard Banner Sizes
From: Tom Hukins
Date: Wed, 11 Mar 1998 15:59:33 GMT

My last post on this topic wandered away from the main point, as James
Tarin and Ben Thompson have pointed out. However, I'm still not convinced
that standard banner sizes are necessary, and I'll try one last time to
explain myself clearly, without wandering away from my main point.

Those who sell advertising space are competing amongst each other for
revenue from those who buy advertising space. Buyers will buy the
advertising space which they expect to fulfil the objective of the campaign
they are buying for at the least cost.

It is highly unlikely that the most effective advertising creative can be
constrained to a small group of standard rectangular sizes. If adverts
could be created in different shapes and sizes, I expect they would perform

Clearly, there are costs associated with creating and customising ad
creative. It would be highly undesirable if we returned to the stage where
different sellers had their own standard sizes which advertisers were
forced to adhere to. This would force advertisers to incur tremendous costs
customising their creative. I am not advocating a return to this situation.

What I am suggesting is that host sites are more flexible with regard to
the advertising they allow. This means that it should be possible for an
advertiser to combine a 40x800 pixel banner running along the side of a
page with a triangular banner at the bottom of the same page.

Technically, it should be possible for host sites to allow such
flexibility. Designers who are treating the Web like yet another graphic
design medium would have problems with this, but good designers would be
able to design pages which allow the incorporation of a wide variety of

So, why isn't this happening?

Host sites aren't flexible enough. Consider the example of the 40x800 and
triangle banners I used earlier. I'm fairly sure that if you phoned up a
selection of Web advertising sellers next week asking to run a campaign
like this they'd laugh at you. Part of the reason for this is that nobody
does it, so sellers don't expect it. Ben Thompson summed up what I expect
would be the main objection in his last post:

>Standardisation means cheapness. If I have to redesign a page for
>every advert displayed I would not have time to do any work (in fact I would
>have to create 60,000+ pages a day by hand).

When Web designers create their pages they do not take into consideration
how they can incorporate a wide variety of advertising into that page. This
is trivial to do. Create a template for the page and process that template
according to the advert(s) being displayed every time the page is
requested. Anyone who really understands HTML and has a rough idea of how
to use Perl, Visual Basic, TCL, or any other programming language can do
this easily. As most sites have a house style applied to all pages, such a
template would only need to be created once, after the site owner has
determined how flexible the site's advertising policy should be.

Sure, there will be a cost to host sites of varying their design to suit
different types of advertising, but the extent to which they experiment is
up to them. Successful host sites will increase their revenue by permitting
a greater range of advertising, without destroying the coherence of their

The Web isn't a WYSIWYG medium. Competent designers are aware of this;
their sites already have the flexibility to adapt to a number of display

Ad designers won't experiment with their creative until they know a
significant number of host sites will accept non-standard ads. Before
de-standardisation can occur, there will have to be a critical mass of host
sites willing to accept non-standard ads. Unless a group of major sites
choose to de-standardise, it is unlikely there will be any move away from
the status quo.

Mat Morrison states:
>Standardisation allows media-owners/audience centres to price and sell
>comparable media.

This statement assumes that we are dealing with comparable media. It isn't
easy to group impressions into groups of comparable media. The same ad will
have different effects on different pages. The same ad will have different
effects on different people.

There isn't even just one standard ad size. 468x60 is, I suspect, the most
frequently purchased, but there are others. The industry seems to cope well
enough with this.

Furthermore, traditional media have coped with this problem. Flicking
through the main section of Wednesday's "The Times" I counted
advertisements of ten different sizes. Some were in colour: some were in
black and white. Some covered less than one eighth of a page: one covered
two pages. In addition, the Personal Column contained several small ads,
which I didn't count. There were supplements for business and sport,
computing and jobs. These sections also contained different sized adverts
placed in different locations.

It is already difficult enough to estimate the value of different
advertising creative. The introduction of non-standard sized ads won't make
this noticeably more difficult.

Test campaigns are a far better solution to any comparability problem than
standardisation. Test campaigns can be used just as well to estimate value
for banners of any size.

De-standardisation would make Web advertising more effective for all involved.

Anyway, I think I've made my point now, so I won't post on this topic any
more (unless anything especially interesting pops up) and risk tempting the
moderator to reject my posts. :)



Learn how to create amazing web sites

Visit eBORcOM's Web Development Resources

[Previous] [Next] - [Index] [Thread Index] - [Next in Thread] [Previous in Thread]