|The "Non-banner" ad|
| 5:26 am on Dec 6, 2000 (gmt 0)|
I just finished the first round of a banner ad comparison test on Yahoo. One was a traditional creative, with a colorful graphic and text (no animation).
The second was what some call a "non-banner". That is, it was designed to look like part of the SERP -- same text as the creative version, but text only, with a grey one-pixel border. Pretty ugly from a design standpoint, or at least pretty dull.
The results blew me away. For 100,000 impressions spread over 6 keywords, the non-banner drew TWICE as many clicks. On one of the kw's it was nearly 5% compared to 1.8% for the creative banner.
Has anyone seen anything similar? This makes me ready to throw out all the creative and go for this style exclusively. I have several theories, but besides the obvious, this banner serves up in a jiffy -- it's only 1.4kb, being simple text and only an 8 color palette.
Maybe more people actually saw it!
| 9:53 am on Dec 6, 2000 (gmt 0)|
this is so very true, it destroys me when you look at the stats. It just doesnt make any sense but a super nice banner, really well designed just doesn't get the click throughs that the basic fake drop down banners get. I find it hard to believe that people click on those fake drop down banners but they do. Movement is also bad, although slap the monkey banner seemed to get everyone in the clicking.
| 10:12 am on Dec 6, 2000 (gmt 0)|
This is a great thread to me because it seems to back up my own theories on advertising.
To put it simply, the closer your advertising is related to the content or function of the page, the higher the click through!
It applies to "fake" banner ads, ads which look like listings, and text links rather than banner ads which jump out at you.
I reject the pessimism of people on the sucess of banner ads. What is happening i think, is that ads that are unrelated to content, that invariably lead to irrelevant offers and scam, are no good, not just banner ads in general.
An effective banner ad looks like the rest of the page, (not that its an external thing) and offers something that a reader of that VERY page wants.
People have been desentised because of the number of ads deleivered by such groups as AdExchange, BannerExchange that are invariably totally unrelated to the reason someone went to the page at the first place.
We dont have hard data, as we havent done any controlled experiments, but since we started using these prinicples out click throughts have imporved massively...
We just accept at the most 12 advertisers throughout our more than 2,000 pages, have only one relevant banner ad on each page (and only on 20% of pages), and text links so advertisers products are fully integrated into the columnists content.
Really interested in reading more on this thread and learning more.
| 10:39 am on Dec 6, 2000 (gmt 0)|
>...the closer your advertising is related to the content or function of the page, the higher the click through!
Marketing types talk about "interruption marketing" vs. "permission marketing". It seems to me that a search Engine ad is in a gray area somewhere between the 2. So many designers forget that and create a pure interruption ad.
Funny thing in this particular test we did -- the worst clickthrough rates came from the keyword that was actually used on the banner! The other kws we bought were synonyms, and they did up to 10 times better. Go figure.
| 10:52 am on Dec 6, 2000 (gmt 0)|
interesting stuff Tedster... seems to spoil my theory!
Wonder if it is possible to be TOO relevant, meaning that it dosent seem to add anything that the other content on the page already offers...
You've got me thinking... Thanks for sharing the stats.
| 11:18 am on Dec 6, 2000 (gmt 0)|
>Wonder if it is possible to be TOO relevant, meaning that it dosent seem to add anything that the other content on the page already offers..
This is only a personal reaction, but now that I think about it, I actually feel a moment of anger when I see my exact search word in an ad. It seems so cheap and manipulative, even invasive. Recently I searched on the word "bula-bula" (Fijian for welcome) and saw promises that I could find "bula-bula" at EBay and Barnes & Noble.
| 2:23 pm on Dec 6, 2000 (gmt 0)|
I always find it amusing when I'm doing a search for a friend's name (John Doe) that although the search engine may not have any results, I can find a WHOLE BOOK about him at Amazon.com or click on his "internet keyword.":)
| 3:35 pm on Dec 6, 2000 (gmt 0)|
>One was a traditional creative, with a colorful graphic and text (no animation).
>The second was what some call a "non-banner".
I believe some of the increase in CTR comes from the fact that the public hasn't trained itself to mentally filter out non-banners. When webzines started injecting ad text in the body (usually at the fold) of the article, they often caught my attention -it took me awhile, but I eventually developed a scan mode that tuned them out, too.
| 10:42 pm on Dec 6, 2000 (gmt 0)|
Yep, these smartypants ad scripts irritate me too and do nothing for the cedibility of the advertiser to me. (personally)
After a search for "Malaysia" I get an offer to "Buy Malaysia" at buy.com and put in a bid for it at some auction site.
It sugessts to me i am at a shopping site and not at a search site that I thought I was at
I immediately change to a search site that doesn't insult my intelligence...
| 2:18 am on Dec 7, 2000 (gmt 0)|
>>Yep, these smartypants ad scripts irritate me too
You've conviced me as much as my data has -- I'm going to avoid using a banner that has the exact search term, whenever I can. I'm thinking synonyms and benefits may be a better way to go than sounding like a parrot.