Building a Baseline

determining causative factors

[chewy]

So I’ve now got 1 day of data from running AdSense.

This is exciting.

Essentially for 4 pages on a site, I am getting about a 10% CTR. Yes I set up channels, but we’re still very early in the game.

I’ve read here that there are things I can do to optimize and I am planning to follow in those footsteps.

My question is this: I read elsewhere here that Google (GAS) may have some sort of optimization process where once ads start getting clicked on Google respiders, reflects and then rejiggers the selection of ads on the site to somehow improve things. I suppose we can call this the “RRR theory”.

Is there such a process?

Meanwhile, if I just left things alone, would I see an improvement? (Given I know I need to average data and that 1 day’s data is only a data point and inferences made on 1 point is like betting on the Yankees or the Red Sox based on the outcome of the first pitch.)

Essentially, is there truth to this RRR theory? What can I expect, generically, over time, if I leave this untouched?

And what kind of time-spans do people find that seem significant?

If I am going to optimize, how do I know that my optimization efforts are the causative factor, rather than Googles RRR’ing the site?

[ken_b]

I have no idea why, but it does seem like targeting can improve as a page ages a bit. That could an illusion though, it might simply be a matter of who is advertizing at the time.

As far as re-evaluating a page (your RRR theory, if I understand what you mean by that), the only way I'm aware of that that might happen, other than Smart Pricing, is when you make a change to the adblock code. Changing the adblock code seems to recall the mediabot fairly quickly. I assume that when the bot returns it also re-evaluates the page content, but others here might have a more deinite answer on that.

How long it takes to get a grip on how Adsense is likely to do on a site, or page, is a matter of time and impressions, I think.

On low traffic pages it can sometimes take a while to get a feel for the result of changes, but not always.

This stuff can be pretty site specific, so there may not be a real "general" rule to follow or use as a guideline.

But I'd say a week is a minimum period to get any kind of idea, a month is better.