New ads work. You will always get a higher CTR
on new ads, both text and graphical (but not Adsense).
Think of it like a honeymoon for ads. New ads are
good for probably 2 to 3 days in terms of decent CTR,
then they drop off to lower levels. I've tracked the
ads from the BlogAds Advertising service for a
period as well. My advice if you're advertising
using BlogAds over a period of time: rotate your ad
graphics maybe once a week.
2. Regular readers tend to be ad blind.
Regular readers will click on ad spots but the CTR
is much lower than for new readers, particularly on
Adsense ads. I've had lots of different traffic in
the last four weeks on a variety of different
subjects and its the posts bringing in completely
new traffic that have the high CTR's, and I'm
talking at times four to five times higher rates on
the same ad spot as per a post that would attract a
3. Old posts = revenue. It really, really
surprised me how many days I'd look at the stats and
see clickthrus from Adsense ads that are so old I
don't even remember writing them! And sometimes the
CTR is higher on these pages as well. Perhaps the
Adsense ad was more relevant that the 2 year old
content? who knows!
4. Niche topics deliver better ad results.
Statistically its niche topics that do best in terms
of CTR, perhaps in part because they bring in new
traffic. Topics that relate to specific occupations,
geographic areas or types of people seem to work
better than general topics relevant to your broader
Top banners aren't very good. I've been playing
with the my top banner space, and it doesn't work
overly well, but it can with rotation. If you're
going to do a banner see if you can rotate it with a
number of ads, alternatively if you've signed the
one advertiser see if they've got a variety of ads
for you to rotate, your CTR rates will be better if
6. Text Links work. Some people might think
that text links are all about Google juice, and
although they might be partially right, they'd also
be partially wrong, because I was amazed to see some
text links in the nav bars doing reasonable CTR
rates. Sure, not amazingly high, but when people
attack O'Reilly and others on the basis of relevance
I'd argue that if one person clicks on it, it must
be relevant to them. In the case on my tracking, it
was a lot more than one person.