Where do user eyeballs go?

[Chris]

Great read:

http://www.poynterextra.org/eyetrack2004/advertising.htm

[abelony]

A great read indeed. I am definitely gonna go back and read this page a couple of more times before building my future websites. I do hate those mouseover ads though. Nothing is more irritating for me than reading an article and with a sligh move of a mouse an ad pops up and covers the part I am reading. Also, they really make the 'Close' or 'Collapse' buttons hard to find sometimes.

[chromate]

Thanks for the link Chris. I've been thinking about doing a similar article of my own for some time. Though, obviously it wouldn't even come close to being as in depth as that one :) There is one point I would like to make though...

They mention the term "eye traffic". Nice term! When considering "eye traffic" the quality of that traffic is just as important as conventional traffic. What I mean is that not only does the user have to look at the ad, they have to be READY to click it otherwise it will get discarded in favour of alternative content. I think this is a really important point to consider.

I think generally a reasonable usage cycle would be:

1) land on page.
2) get some sort of satisfaction from the page's content (did the user find what they came to the page for?)
3) look for an "interesting" exit point
4) leave page.

For example, the EyeTrack results show that there is a general hotspot in initial viewing that falls within the top left area of the screen. Now, I would suggest that if an ad was placed here its effectiveness would be diminished because it doesn't fit in with the cycle above. The user would have to skip point 2, which is the reason for the user being at the page in the first place. So they would be more reluctant to click out. If the ad is amongst, or to the right of, the content then the ad is naturally at a point where the user will be looking for their next "click" and so it may perform better.

It's not just enough to have the traffic. The quality of the (eye) traffic is just as important.

[Jaffro]

Without actually reading the article i've always tried to consider this problem that chromate talks of. Supply the visitors with what they want, but not providing them with everything so that they look for an 'exit point' on your site. For SEO sites if your site provides nothing but ad's the user is more likely to just click back and look down the search list, but if you give them good worthwhile reads they are more likely to look further around your site for a good exit point - instead of moving back to the se results.

I dunno - got work to do :) Just wanted to say i agree with chromate on the satisfaction and exit point things.

[Chris]

Don't forget about impulse clicks. The ad might not further the user's interest on the topic they originally came to your site for, but it might create an impulse the user did not know he had.

[ozgression]

There is a great blog post about "users on a mission" at http://themanfromscandinavia.blogspot.com/2004/09/mattias-take-on-future-of-online.html

FWIW, I've found placing advertising at the bottom of articles etc. performs better as they have finished finding what they need and are ready to go somewhere else.
________
DLD engine (http://www.ford-wiki.com/wiki/Ford_DLD_engine)

[Westech]

I wholeheartedly agree with Oz on this one. Placing adsense ads at the bottom of an article while suggesting that the user research the topic further for himself has worked out well for me. For example, if the article is about compact cars and I know that the adsense ads are going to be from car manufacturers, at the end of the article I'll say something like "now that you know what to look for while shopping for a compact car, you should do some research at the various car manufacturers' websites to find the car that best meets your needs." Right below that, I'll put the adsense block, blended in with the content of the website as much as possible. This way, at the end of the article when the visitor goes into "what should I do next mode," I'm telling them what they should do next, and providing a convenient way to do it.

[nohaber]

The article is very good.
With the recent changes in Adsense Tos, we can place ads at the top and at the bottom of the article :) It would be interesting to track actual clicks on pages with multiple ads. Where do visitors look first, what do they read next etc. and if an ad was clicked which of the ad panels was it in?

[Dan Morgan]

That downward trend to 4-5 clicks per 1000 ad impressions (0.4/0.5% CTR) - I wonder if that is an average across the formats (i.e. high 20% CTR on Eyeblasters - through to 0.05 on some 468x60 placements?

[incka]

Perhaps that is because doubleclick has boring ads?

My CTR on leaderboards is about 2.5%...

[emprivo]

Great read:

http://www.poynterextra.org/eyetrack2004/advertising.htm

Chris and everyone else who was interested in this article, I'd highly recommend "Don't Make Me Think!" by Steve Krug. One the best books I've read on usability -- and one not written by self-proclaimed "experts" like Jakob Nielsen.

No offense to people who adore Mr. Nielsen.

This book was informative, without the fluff, and ridiculously funny every step of the way. You won't feel bored for even a second reading it.

[chromate]

I've heard about this book and skimmed over the first chapter on amazon. Looks pretty good.

[LuckyShima]

No offense to people who adore Mr. Nielsen.

Nielsen certainly was a guru in his day, but it wasn't that difficult for him because there was no competition.

Also, I like the results presented in the article on this page:

http://www.poynterextra.org/eyetrack2004/main.htm

which show what a hotspot the top left corner is. It was always a bit of common sense to use this corner for branding and this supports that view.

[Dan Morgan]

Perhaps that is because doubleclick has boring ads?

My CTR on leaderboards is about 2.5%...

Doubleclick adserve to a wide variety of publishers, thats it - they do not have a media network anymore.

Also, I assume they are taking about sales performance ad campaigns from bluechip companies/household names rather than the Smiley Central/Bonzi/1,000,000th visitor, Free Holiday Vouchers etc where compelling (or annoying ;)) ads are the norm.

Dan