Welcome to WebmasterWorld Guest from 23.23.36.47 register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts, Become a Pro Member
Forum Library, Charter, Moderators: buckworks & eWhisper & skibum

 Tweet
CTR vs. conversions - a Catch-22
So just how much does CTR affect CPC?
limitup

Msg#: 3424299 posted 1:13 am on Aug 17, 2007 (gmt 0)

Here's my dilemma ...

Ad #1 has a CTR of 4.68%
Ad #2 has a CTR of 3.88%

Ad #1 has a conversion rate of 1.7%
Ad #2 has a conversion rate of 2.0%

Ignoring the CTR/CPC relationship, obviously I want to continue running Ad #2 because it would yield a higher ROI.

But just how much does CTR affect CPC anyway?

I'm just delving into this stuff, but it definitely doesn't seem to be linear because in this example my CPC is exactly the same with both ads.

Perhaps a higher CTR will give you a lower CPC up till a certain point, but then it doesn't matter?

In this case are my CTRs so relatively high that I'm no longer getting any CPC benefit, regardless if my CTR is 3.88% or 4.68%?

RhinoFish

Msg#: 3424299 posted 1:45 pm on Aug 17, 2007 (gmt 0)

to help you, we need to view the data set's sample size. it's a common error to read too much differentiation into small differences in stats from too small a dataset.

if your ads are substantially similar in content and meaning, which is often the case, can you explain the real life difference in the results you're seeing? there are legitimate answers when the ads are sufficiently different, but that's often not the case.

limitup

Msg#: 3424299 posted 6:14 pm on Aug 17, 2007 (gmt 0)

Thanks for your reply. These stats are based on 1000s of clicks. The ads are "substantially similar" however Ad #1 uses the word Free in the title which obviously generates a few more curiosity/tire kicker clicks. That is my explanation for the lower conversion rate.

SanDiego Art

Msg#: 3424299 posted 8:46 pm on Aug 17, 2007 (gmt 0)

How many keywords are these ads showing for?

Sometimes the stats can get washed out because Ad #1 could have a 9% CTR on term1 and 1% on term2.

You might want to run an ad performance report to see if there are any major differences for the ad copy PER term.

If you do have a lot of keywords sharing the ads, try breaking it down to smaller groups to see if the ad copy performs differently.

Also, if ROI more important than volume - then yes Ad #2 is better. But Ad #1 actually producing slightly more conversions per impression.

Ad #1 : 10,000imp x 4.68% x 1.7% = 7.956 conversions
Ad #2 : 10,000imp x 3.88% x 2.0% = 7.76 conversions

So for a volume scenario: Ad #2

thecloser

Msg#: 3424299 posted 8:52 pm on Aug 17, 2007 (gmt 0)

You're still firmly within the margin of error. Given how close the results are thus far, you're likely looking at thousands more clicks before you have a conclusive result, beyond the margin of error.

RhinoFish

Msg#: 3424299 posted 4:00 pm on Aug 18, 2007 (gmt 0)

i concur with more data being needed and with segmenting things that perform well further (they're worth the time of managing them more finely). but more generally, i still can't answer the question fully without more data - my instincts tell me your CTR is still too low and you're proposing a choice between two things that aren't yet at their full potential, when you should be chasing a higher ctr. but given that i don't know the specifics of what you're bidding on and the avg positions you're occupying, i could be wrong - you might be rocking already. So, without knowing for sure, i'd suggest you pursue more ads and more segmentation to achieve higher ctr still - you're in the average range for pro performance (which as i said is a generlaiztion and simplification and i could be off here depending on what keywords, concepts and positions are), and you need to do better on the ads.

since one says free and one does not (if it were really free, all your ads would say free), i'm imagining that free isn't exactly true - otherwise it wouldn't attract the disinterested - it's either a good deal to be highlighted (and conversions go up) or it's not quite the deal its portrayed as, in the ad... make an ad that describes the offer as best you can and conversions will be better. make an offer that stirs people to buy - do something different than others.

mike_ppc

Msg#: 3424299 posted 12:03 pm on Aug 27, 2007 (gmt 0)

Sorry, San Diego, but I think it's the other way around: no.2 for ROI and no.1 for volume (smaller ROI as you are paying a lot more clicks). Anyone else?

onlineleben

Msg#: 3424299 posted 1:53 pm on Aug 27, 2007 (gmt 0)

 Ad #1 : 10,000imp x 4.68% x 1.7% = 7.956 conversions Ad #2 : 10,000imp x 3.88% x 2.0% = 7.76 conversions

If the differential in CTR and CR were higher, the discussion would make more sense. With the Example of 10,000 ad impressions for each ad and the above mentioned rates for clicktrough and conversion, both ads result in 8 conversions (a fragmented conversion is only of statistical value).
To feel a real difference of which ad is better, you need at least 100,000 imps per ad to have a difference of 2 (79.56 -> 80 and 77.60 -> 78).
Regarding ROI ad #2 is better in your current scenario as CPC is the same for both ads and you need fewer clicks to convert the same number. When switching to the 2x100,000 imps scenario you have to figure out if the cost of 800 additional clicks you would receive from ad #1 is covered by your increased conversions of two units.

 Global Options: top home search open messages active posts  Tweet