Were you proactive?

[davewray]

Hi all. I'm not sure why I waited so long, but a couple of days ago I went from a banner format for Adsense on my main page to an inline rectangle format.

The next time I checked my stats I was shocked! It's been two days and my site gets a statistically significant number of clicks as it is. Well, my CTR more than doubled! I could not believe it!

Well, now I'm worried that my spike in CTR will flag Google and may draw unnecessary attention to my site. When I got the "tax info" email from Google yesterday my heart jumped thinking it was "the" email.

Now, my question is this. When you guys made a change to the layout of your Adsense which resulted in a big upwards spike in CTR, did you contact Google to let them know why, or did you sit cool? I'm thinking an email to G explaining why my CTR has spiked so high would be good, but I'm not too sure. TIA.

Dave.

[401khelp]

did you contact Google to let them know why

Congratulation on the improved CTR!

When we made a similar change that also resulted in an improvement in our CTR, we did not contact Google. Figured that as long as it continued at the new CTR, Google would see that as just the new baseline. Spikes can happen for any number of reason other than fraud. For example, a new advertiser with a compelling ad that draws an increase number of clicks.

[richmondsteve]

Google knows what ad formats you're using via the JS code and I have no doubt that their fraud detection algorithms take observed program-wide ad format averages into account. Don't contact them in advance - there's no need to flood their bots and staff with unnecessary emails that just drive up their support costs.

I know plenty of publishers who switched formats and had an immediate bump in CTR (myself included) and none got any kind of warning email as a result. Relax.

[alika]

We changed from banner to leaderboard in our homepage and CTR increased by 35%.

No, we did not contact Google. We are not paranoid enough to contact Google as we trust Google to know that the spike is not a result of fraudulent clicks. We changed the format around 2nd week of October and no email from G.

[davewray]

Thanks guys. I normally wouldn't be so worried, but I'm talking about a pretty big CTR and that's what I'm concerned about! But, I'll trust that Google can see what changes I've made and look at industry-wide averages to make informed decisions.

Dave.

[davewray]

Alika...A 35% spike I can understand you not getting paranoid about. But what about a 100% spike? And this is after tens of thousands of impressions....

[alika]

On the first two days of the change, the spike was around 175%. The first day change alone hit around 300%. Then CTR gradually stabilized and decreased. The 35% was the average for the nearly 3 weeks.

I don't look at 1 day spikes. I look at 7 day averages at the very least.

[alika]

correction - the 175% was the first 5 (and not 2 days).

[JohnKelly]

Dave,

Went from skyscraper to inline rectangle back in Sept, and saw a 4x increase in CTR. I did not notify Google, just sat back and enjoyed the increased revenues.

While Google may possibly investigate the increase in your account's CTR, since you did nothing wrong they should not find any fraudulant clicks. Google expects some experimentation from publishers to maximize revenue.

Congrats on the increase, it's amazing sometimes what a small change can do.

[esllou]

i changed something three weeks ago and have had a 300% increase in CTR ever since. I had a major shock when I checked my stats after the first day and DID contact google. It was being paranoid, it was being over paranoid, but I did it anyway. NOT because it was adsense and not because of THE e-mail that people are worried about getting.

I just think it is good affiliate relationship manners. I did the same to allposters when I tinkered with something and doubled sales overnight. I let them know, also as a way of educating them as to increasing sales.

[Blue_Fin]

Everyone who has gone from the 2 ad banner to any of the 4 ad formats has reported significant increases in CTR. This started on August 21st when Google implemented the new ad formats. If you scroll through the posts back until that time, you'll see several topics that address this. No one has reported hearing from Google after implementing this change.

[davewray]

Blue_Fin.....which makes me wonder why it took me SO long to change formats! :)

Dave.

[daunk]

Just changed my site and will report back on the change in CTR etc... could be interesting as my site gets about 16k impressions a day so even a .2% increase would make a decent enough difference. Have to say that if adsense was CPM its about $1/8. Used to be more but its decreased :(

[cyberprosper]

My first "the email" was in fact after I switched formats and my earnings/click thrus almost doubled.

[Jolly_Roger]

I was getting good performance from the skyscraper (integrated colors, high on the left and snuggled close to my main content), but I have never tried the leaderboard. Today I switched my site to use the leaderboard. The site doesn't look as good to me now and doesn't scale down to narrow browser windows -- but I am going to decide whether to keep it based on numbers. ;)

Going from a top-of-page banner to a top leaderboard is probably a lot more predictable than going from a skyscraper to a leaderboard. I've seen several comments here about that change but not with accompanying ad location and color information. Unfortunately for the usefulness of my data sample, I also had to do a bit of a redesign in order to fit the leaderboard in -- so my info will be compromised as well. I really have no idea what to expect.

However, I do expect any changes in ctr to be mostly the effect of the leaderboard. Of course, if page-views-per-visitor drops like a stone from the redesign I'll have to bail out on my changes and go back to the skyscraper and will learn very little.

Anyway, in a belated effort to be proactive, I feel I had to try it. Some time early next week I should have an idea about the performance difference for any of you who are interested.

Meanwhile, does anyone else have experience in going from a high left skyscraper to a leaderboard (or vice versa)?

[Jolly_Roger]

I aborted my (first) leaderboard test after only a few thousand impressions covering less than a day. My CTR dropped by as much as half, depending on the adsense reporting delay. However, the layout change resulted in a larger number of page views per visitor which, of course, accounted for some (perhaps a third) of the CTR drop.

My plan now is to let the new format stay up -- with the skyscraper -- until the numbers stabilize. Then, depending on what I learn from that, I may try the leaderboard again. If the new format results are consistent with my current assumptions, however, I won't bother put the leaderboard back up.

From what I've seen so far, however, I think the wide variability in possible placement of the skyscraper accounts for the lack of pattern we've seen here in the few reports of skyscraper vs. leaderboard performance. That is, comparing a below-the-fold skyscraper to a top and centered leaderboard will surely show the leaderboard performing best. A high top left skyscraper will, of course, be more competitive with the leaderboard -- and, as my early but flawed results suggest, may beat the leaderboard substantially.

One other factor that may work in favor of the skyscraper at my site is this: my site has almost no graphics (only a text-looking gif logo) and has no other ads other than adsense. Therefore, going from a well-integrated skyscraper to a leaderboard makes a rather jarring difference in how 'commercial' my site looks. At first glance with the skyscraper it looks ad-free. At first glance with the leaderboard, well, it looks like a giant ad front-and-center. YMMV.