My little experiment

Effect of high traffic site over an established one

[santah]

So here's the setup.

I have small amount of pages (5-6) on this one site, generating about 5k adsense impressions and making $X per day, for more than 5 years now.

The traffic is pretty stable, so are the earnings.

That's all I have in my adsense account.

Recently, I decided to see how one of my higher traffic sites will perform, added adsense code to each page (split into channels for each section) and turned it on.

The result - 50k new impressions per day - the earnings - + around 5%-6% (and the overall earnings of my older channels dropped).

I monitored this for 2 days, and turned the ads on the new site off.

This resulted in $X/2 earnings for the first day that I had only my old ads in my adsense account (with the same constant traffic I've had forever), and on day 2 - it returned back to normal (where it is ever since).

So yea, I don't want to damage the great performance of my old ads with my new ones, but I wanna try and monetize my high traffic site with adsense.

Having multiple adsense publisher IDs would've been best I guess, but since I don't - what do you guys suggest? What's the best path to take here?

Cheers

[DaStarBuG]

Maybe you should run AdSense on your new site for about a month to make a valid decision.
It takes more then 2 days for AdSense to deliver targeted Ads to a new content rich website.
And who says that the return to normal wouldn't have happened to your old website if you had left the AdSense code on your new website?

If I run experiments I run them for at least 2 weeks (bare minimum) but most of my experiments run for 1+ month.

[santah]

Probably will at some point.

But for now, I'm convinced the new ads will improve with time, while my old ads won't get back to the performance they had. Smart pricing, or whatever it's called nowdays won't let an ad perform as well in an environment with x10 the traffic on the same adsense account (even though they're on different sites, adsense channels etc.).

[netmeg]

Sure it will. But you're just going to have to trust me on that, because I'm not going into specifics.

[Leonard0]

Interesting experiment.
Are both sites about the same topic?
If so, maybe you are experiencing burn-out of advertisers' spend as theorized in this thread a few days ago:
[webmasterworld.com...]

Is it possible to increase the page impressions gradually, say by 1,000 impressions a day?
You're probably triggering some sort of throttling effect by increasing impressions by 10x.

But Netmeg says that it will eventually improve, so it probably will, no matter how you do it.
Unless, of course, there really is a cap? :)

[netmeg]

There's no cap or ceiling, except whatever your niche inherently has.

[santah]

Not the same topic. Totally unrelated topics and themes of the sites.

And yea, I'll probably try it again in a few days (when I finish testing some other ads) and will report my findings in here.

[santah]

Just got an idea.

What if I monetize the new site through google's DFP instead of directly through adsense?

Will test and let you know :)

[DaStarBuG]

DFP is an Ad Server not an Advertising Network like AdSense.
You can utilize AdSense through DFP however there will be no difference in earnings between AdSense via DFP and AdSense only.

[dataguy]

So are you saying that with more ad impressions, you earned less money overall?

[santah]

No, I earned about 5% more, but my old ads started performing way worse (and still did for about a day after I turned the high volume traffic off).

Btw running adsense through DFP seems to be working a bit differently than pure adsense (new traffic doesn't seem to be hurting my old ads). I'll post more after monitoring it for a week or two.

[IanCP]

As others may have commented, it's very dangerous drawing conclusions after making any changes.

All too often a positive or adverse affect "coincidently" ensues which has absolutely nothing to do with those changes.

I think most of us have been embarrassed there at some time.

[ADDED] Often it takes at least a month to draw any conclusions.

[dataguy]

I've seen it where I had many more ad impressions and I earned much less overall, but I've had a hard time finding anyone who can duplicate my findings.

[netmeg]

I definitely had a lot more ad impressions running through DFP, but no effect on earnings. That was a few years ago, I'll be firing it up again to test later in the year.