Click Through Rate On Adwords

[bathgems]

Should I care if one variation of an ad has a 5.5% CTR and another slight variation has a 4.5% CTR? I mean, I pay the same CPC either way, right?

I guess Google's argument would be that I get higher placement with a higher CTR but in the end my ad is constrained by my daily budget anyways.

Thoughts?

Steve

[MattE]

your cost per click may not be the same - check that in your account area.

The end of the day is the conversion rate - a 0.5% clickthrough with a 10% conversion rate is better than a 5.5% click rate and a 0.5% conversion rate. Why the difference in conversion rate - expectations based on the ads and if you have different landing pages.

[arvana]

Also if you have variations on an ad, Google will automatically place the one with the highest CTR most often (and possibly higher as well).

[MattE]

Also if you have variations on an ad, Google will automatically place the one with the highest CTR most often (and possibly higher as well).

I believe this is only if you have that function enabled.

[bathgems]

So the only issue is the conversion rate of the ad? The click-through-rate is otherwise irrelevant?

Unfortunately, my items do not seem to show up in my conversions in analytics most of the time. Between customers who want individualized quotes where I send an invoice through Paypal or the website being passed among several decision makers prior to purchase, my sales do not seem to correlate to my tracking.

Steve

[SearchEngineZ]

There is a direct relationship between CTR and cost. If you manage to double your CTR by writing a better ad, your cost per click will halve.

In general, if the ads convert to sales at the same rate, it is best to delete the ads with the poorest CTR.

Before deleting an ad with poor CTR, consider:

1) If conversion to sales of individual ads is not being tracked, could it be that the best ad might have a higher CTR because it is attracting clicks from folk who aren't likely to purchase? Having the word "free" in the title can do that.

2) Is the CTR report from a long period, like a month or more? The longer the period, the more accurately the CTR will predict future CTR.

3) Have there been enough clicks for the CTR to be statistically accurate? For example, would it take just a couple of extra clicks for the poor ad to equal the best ad?

[tomfoolery]

Gotta agree with SearchEngineZ. Over the last 18 months, I have been slowly improving my ads to boost CTR. As a result, there has been a significant drop in my cost-per-click (and thus my overall cost). My average CTR went from 3.5 to 6.8 (ish), and I've seen about about a 50% reduction in CPC. This is massive for my business (and for the clients that I manage -- some on a limited budget)

That said, the impact between a 5.5 and 4.5 may be massive to your CPC, or insignificant - it depends on what kind of traffic you're talking about. If the difference is due to a handful of clicks vs thousands of clicks, then you're comparing apples to oranges.

[deepsand]

Originally Posted by MattE

I believe this is only if you have that function enabled.

1) The default is for the variations to be displayed at a rate dependant on their respective CTRs; and,

2) Changing the setting for such does not yield a true A-B split, as should occur. In fact, the actual result can vary greatly from the ideal; I've seen 75%-25% ratios!

[deepsand]

Originally Posted by bathgems

So the only issue is the conversion rate of the ad? The click-through-rate is otherwise irrelevant?

Unfortunately, my items do not seem to show up in my conversions in analytics most of the time. Between customers who want individualized quotes where I send an invoice through Paypal or the website being passed among several decision makers prior to purchase, my sales do not seem to correlate to my tracking.

Steve

1) Include in your ad variations some distinctive element such that you can ask respondents which variation they responded to.

2) Armed with that data, you can then calculate your Cost per Conversion for each ad.

3) Keep record of the Sales Revenues generated by each ad.

4) With this data you have all you need to calculate your Revenue per Click and your ROI.

[craigmn3]

I read somewhere, perhaps here, that google will religate poorly performing ads to the side listings and take them off the top results. I guess they want their money, though I tend to disregard the top PPC in favor of the side. But I prefer the organic listings, figuring anyone who is paying per click is going to be charging me in the price of their goods.....though that is not always or even often the case. Just psychology I guess

Craig,
Web Geek
Edited by mod Webnauts. You may have links to your web site only in your signature.

[deepsand]

[quote="craigmn3"]I read somewhere, perhaps here, that google will religate poorly performing ads to the side listings and take them off the top results. I guess they want their money, though I tend to disregard the top PPC in favor of the side. But I prefer the organic listings, figuring anyone who is paying per click is going to be charging me in the price of their goods.....though that is not always or even often the case. Just psychology I guess

Craig,
Web Geek
Edited by mod Webnauts. You may have links to your web site only in your signature.

Those who achieve high placement in the organic listings do so at a real & measurable cost.

So, there's no merit in the belief that those who choose to use paid advertising by necessity have higher costs, and therefore higher prices, than others.

[bathgems]

Plus what is more likely to bring you relevant sites, an abstract algorithm or the self-selection of someone putting their money on the line?

I tend to think the paid listings are more relevant.

Steve

[incrediblehelp]

Originally Posted by bathgems

I tend to think the paid listings are more relevant.

Maybe in some instances, but the fact that they are "paid" means bias is involved. The more you pay the higher ranking you get, meaning to me that relevancy is lost.

[deepsand]

Originally Posted by incrediblehelp

Maybe in some instances, but the fact that they are "paid" means bias is involved. The more you pay the higher ranking you get, meaning to me that relevancy is lost.

1) Organic listings are not without bias.

2) Over-spending on PPC listings is ultimately self-correcting.

[incrediblehelp]

Originally Posted by deepsand

Organic listings are not without bias.

Of course, someone has to create and control the algo, but that is done by the SE themselves not each of us by bidding higher.

[deepsand]

Originally Posted by incrediblehelp

Of course, someone has to create and control the algo, but that is done by the SE themselves not each of us by bidding higher.

How or why bias is introduced is irrelevant. Furthermore, it is unavoidable, given that there is no common goal among all searchers.

Google, for example, is biased in favor of those who seek information about a good or service, as opposed to those who seek a source of such good or service.

[incrediblehelp]

I think I am being taken out of context now. I simply saying relevancy in paid listings is not as good as it should be in organic listings. Paid listings are placed by bid amount with some small, small relevancy factors, while organic listings are all supposed to be based on relevancy alone.

[deepsand]

Originally Posted by incrediblehelp

I think I am being taken out of context now. I simply saying relevancy in paid listings is not as good as it should be in organic listings. Paid listings are placed by bid amount with some small, small relevancy factors, while organic listings are all supposed to be based on relevancy alone.

The publishers of PPC listings, i.e. Google, Overture, etal. are responsible for the relevancy of both the organic and PPC listings.

In my experience, they do a better job, re. relevancy, with the PPC listings, in two ways:

1) PPC listings are open to review by the publisher prior to their being published; and,

2)The publishers will review an irrelevant PPC listing if reported to them. Trying to get an irrelevant organic listing reviewed, let alone un-indexed, is next to impossible.