CTR direct relationshp with eCPM?

[Raymond]

I notice when I have high CTR, eCPM goes up, and when CTR goes down, eCPM goes down with it. I have been playing around with the placement of this ad to create a fluctuation in CTR. The % increase of eCPM is quite huge.

Without giving the exact stats, please use your imagination on these numbers: (I thought it is against TOS)

2.7 - 7
3 - 7.1
2.2 - 5.4
0.6 - 1.3
0.88 - 1.4

These numbers are from 5 different days of over 700,000 impressions in total. Does anyone else see the same pattern?

I am quite new to adsense. I apologize if this is a common knowledge.

[jimbeetle]

There could be a corelation, though I'd hazzard a guess that it's not the fact that CTR goes up, but maybe the underlying cause that makes the CTR rise.

Good well-written ads, properly targeted should cause more interest among your site visitors, therefor more clicks on possibly higher paying ads.

Conversely, if there happen to be periods of low quality ad inventory, CTR is likely to go down and those clicks might be on lower paying ads.

Or, it could be any number of other factors that affect CTR.

[angel1949]

jimbeetle means the correlation measure affected CTR and eCPM? :)

[hunderdown]

Raymond, there is indeed a direct relationship between CTR and eCPM.

eCPM is calculated like this, with both your EPC and your CTR written as decimals:

CTR x EPC x 1000 = eCPM

So if your CTR doubles, and EPC stays the same, your eCPM will double too.

Here's an example, using a 5% CTR and a 20 cent EPC:

.05 x .2 x 1000 = an eCPM of $10

Play around a bit with that and you'll see how it works.

[jomaxx]

eCPM is basically CTR times EPC. Of course if you double your CTR your eCPM will also double (on average, more or less).

If those numbers are your CTR and your eCPM, then the relationship is actually pretty steady.

[Raymond]

Ok I understand now. Thank you so much for the clarifications.

[jimbeetle]

jimbeetle means the correlation measure affected CTR and eCPM? :)

What I was trying to say is that instead of looking at it strictly as a formula, that there could be other underlying causes.

jomaxx basically stated if your CTR doubles and your EPC stays the same your eCPM will also double. Makes sense. But the only time CTR directly affects eCPM is when the EPC stays the same.

I'm not saying what I want to state too well. I guess I just think there are so many factors that affect the CTR and EPC I think it's too simplistic to say that a rise in CTR always equals a rise in eCPM.

[angel1949]

thank you~

[Pengi]

jomaxx basically stated if your CTR doubles and your EPC stays the same your eCPM will also double. Makes sense. But the only time CTR directly affects eCPM is when the EPC stays the same.

This is just not true.

eCPM is calculated and changes in CTR will always directly relate to changes in eCPM - but there may be other changes as well.

Think it through the real (hard) data are:
1) Total earnings (made up as the sum of the earnings from each click)
2) The total number of clicks.
3) the total number of impressions.

CTR is calculated as 2) divided by 3) [this isn't an opinion, it is a fact]

eCPM is calculated as {1) * 1000} divided by 3) [again, this isn't an opinion, it is a fact]

EPC is calculated as 1) divided by 2) [again, this isn't an opinion, it is a fact]

eCPM = {total impressions * CTR * EPC * 1000} divided by {total impressions}

eCPM = CTR * EPC * 1000 [fact based on mathematics]

[jimbeetle]

I know the formula and I know how to do the math (I'm an old math textbook writer and editor). But I'm trying to get away from a purely formulaic response. And I guess I again didn't state what I was trying to say as clearly as I wanted.

Raymond:

when I have high CTR, eCPM goes up, and when CTR goes down, eCPM

That can be true in many cases, but it can as easily be false as it excludes the EPC. CTR can go up and EPC go down, resulting in a lower eCPM. So, a rise in CTR does not automatically mean a rise in eCPM.

jomaxx:

Of course if you double your CTR your eCPM will also double (on average, more or less).

Only if the EPC stays close to the "average, more or less".

I was trying to get to the underlying reason for the fluctuations in CTR and how that might affect the eCPM. Raymond's CTR for the few days he used ranges from a high of 3.0% down to 0.6%. Assuming that he's seen these wide variations quite often you have to ask why.

[jomaxx]

I was trying to give a simple answer to a simple question.

We could have a hundred-post thread on possible scenarios and postulated relationships between CTR and EPC in the AdSense program, but I doubt it would help the OP.

[jimbeetle]

We could have a hundred-post thread on possible scenarios and postulated relationships

Yep, that's why I gave one example in my first reply. It's much more than a "if CTR goes up eCPM goes up" correlation and the are too many other variable to consider.

[Pengi]

I was trying to give a simple answer to a simple question.

Absolutely. And you were spot on. eCPM does equal CTR * EPC * 1000.

This is not an opinion or a guess. It is the answer! No one is saying that eCPM only depends on CTR (it also depends on EPC) but no one can deny that, if everything else remains constant, any change in CTR will relate directly to a change in eCPM.

[danimal]

great thread.

raymond, when you are doing ad placement, sometimes driving up the ctr can result in lower epc, later down the line.

a worst-case scenario for that would be a page with no outgoing links except for adsense, which is never a good idea... or perhaps a page with all the outgoing links beneath the fold would encourage lazy bored clickers who just want out, and have no real interest in the products they are clicking on.