Question about Tracking Conversions & Picking Winners...

Status
Not open for further replies.
B

blacktastic

Whats blacker than black?
Dec 13, 2006
66
0
0
On the web (and in shadows)
Okay, so I've been trying to do better about lowering the number of keywords I have in adgroups, which has allowed me to also create more relevant ads for each adgroup.

However, when I try to decide which ad is performing better, it would take me months in some adgroups to grab enough conversions for so I could be statistically sure that one ad is better than another for an adgroup.

Not as much a problem for my aff sites, but my PPC clients aren't as patient. Any ideas? Also, how many conversions do you all wait on before deciding which ad is better? I've heard everything from 17 up to 100.
 


As far as your other problem, I would say that you may be possibly looking at it wrong. You will want to use the conversions to determine which keywords are performing the best, so that you can manipulate the CPC on those. You will use the adgroups' number of clicks to determine which ad is performing better, replacing the lesser performing ad (the one with the lowest CTR) with a better performing ad, eventually coming out with the most superior ad for the keywords.

So in short, use the clicks per ad to determine the better ad, and replace it. This shouldn't take too long, because you aren't waiting on a conversion, only clicks, and you usually only need 30 or so to find out which is best.

Then to decide which keywords or adgroups to keep, compare the total commission received on a particular keyword to the spend on clicks. Keep whatever is profitable, lose whatever isn't profitable. The general rule of thumb on this is to spend double what a single conversion would be, and if it hasn't received any conversions, dump it.
 
I can definitely see what you're saying, but I've had cases where I wrote killer ads in terms of a great CTR, but it drew a bunch of spaz-clickers who didn't convert.

In those cases it wouldn't be the keyword's fault it didn't convert - I just sent less qualified traffic.
 
You don't necessarily want killer ads that attract clicks, you want killer ads that attract customers.

If you are in a market where you want to attract as much traffic as possible, killer ads that attract clicks may be fine. If you are in a specific niche, you may need to 'weed out' bad clicks by making your ad attractive to your clients, while repelling tire kickers.
 
Think of it this way, say my product is a Real Estate product that costs the user $20 for shipping and handling.

Here's an ad that will attract clicks, but not necessarily conversions:

Make $1000s Weekly
Make $1000s Weekly In Real Estate!
Financial Freedom Around the Corner.

This one may weed out some of the tire kickers:

Make $1000s Weekly
Make $1000s Weekly In Real Estate!
Low Cost. Pay Only Shipping.

The second one lets the user know there may be some cost, so you get more qualified buyers. Its not always a good approach to repel clicks, but sometimes its needed, depending on what you're selling. Some markets you need a sniper rifle style ad, to get the most qualified buyers, but in some markets, hand grenade style ads will work, attracting as many eyes as possible.
 
for sure - some of the methods for ad writing for better conversions are obvious, but take some of the examples you see over on mindvalleylabs' blog. What about if one is past the basic testing and is tweaking to play with the big dogs in a niche? Even seemingly insignificant changes can have substantial impact which could add the need for ad tracking in addition to keyword. w/ client work where I have access to thank-you pages, that is a breeze but with my affiliate tracking system I don't have ad tracking available.
 
Status
Not open for further replies.
Top Bottom
Back