PPC > CPA Questions

Status
Not open for further replies.
J

Josh P

New member
Aug 11, 2006
169
3
0
Ok, here's the deal. Looking into doing PPC to CPA, so I set up a site, got my adwords campaign plugging away... here are stats for that:
45 clicks, 5.03%CTR (this is very good)
So my problem isn't getting people to my site.
Once they are one my site, which only has links internally (like 2) and links to the CPA offer. Close to 50% of my visitors are clicking the link to the CPA offer, but of those, only about 9% are actually doing the CPA... so I'm looking at 4-5% overall conversion rate = mass suckiness.
I don't get it though, my ads are great, my CTR TO the offer is great, but then the people who complete it totally drop off. I'd love to show one of the people who know these things (Chrisingle, engaged, Dru, kaveman, etc... :bowdown:) my site, I really need an honest critique on this. the EPC I have now is like..4-5 times lower than the average total EPC for the ad network, which isn't good. hit me up if you're willing to help out, I'd greatly appreciate it.
J
 


Well it's hard to say, without knowing how many keywords your using, what keywords your targeting and how they relate to your ad/cpa. Plus maybe 45 clicks wasn't a long enough test.

Anyhow, you say 50% of your traffic is doing nothing, so your basically throwing half your money away right there. I would first ask why? Could be the keywords you've chosen needs to be narrowed down or maybe your ads are not truly reflective of your cpa offer.

Make sure your ads a not to vague or misleading. If this is not it, then maybe it's a keyword issue. Find and use only the ones that and doing good and reflect them in your ad and landing page titles.

You could also try collecting emails and do follow ups on them afterwards.
 
Josh P said:
Ok, here's the deal. Looking into doing PPC to CPA, so I set up a site, got my adwords campaign plugging away... here are stats for that:
45 clicks, 5.03%CTR (this is very good)
So my problem isn't getting people to my site.
Once they are one my site, which only has links internally (like 2) and links to the CPA offer. Close to 50% of my visitors are clicking the link to the CPA offer, but of those, only about 9% are actually doing the CPA... so I'm looking at 4-5% overall conversion rate = mass suckiness.
I don't get it though, my ads are great, my CTR TO the offer is great, but then the people who complete it totally drop off. I'd love to show one of the people who know these things (Chrisingle, engaged, Dru, kaveman, etc... :bowdown:) my site, I really need an honest critique on this. the EPC I have now is like..4-5 times lower than the average total EPC for the ad network, which isn't good. hit me up if you're willing to help out, I'd greatly appreciate it.
J
Click to expand...

You have to be more specific. Is it a zip/email based offer or does the visitor have to fillout a form?

And if your your epc is so poor maybe target more specific keywords.
 
Thanks for the replies (and for other people helping me) but I've got it turning profit now :)
 
I totally scrapped that campaign, and started a new one (different offers, different landing page). Offers with less competition.
 
Sounds like you were on to something, but you scrapped it too soon. You only ran the campaign for 45 clicks which is no where near enough to generate statistically significant data. I wouldn't worry about the 50% initial landing page bounce rate, I have campaigns with 50-60% BRs and are super profitable. I'd keep at at, testing and tweaking.
 
bad offer

Could the offer just suck? It seems I always have to try a handful of offers before I find one that really converts.
 
yes, that offer did just suck.
and Sonic, 45 clicks is definitely enough statistical data. More than 20 data points is accepted in the world of statistics. And yea, I was on to something, I took what worked of the campaign, and changed some things, and now I have a working campaign.
 
30 data points is typically the minimum required in statistics, and that is when comparing multiple test to determine a winner and even then your confidence level is going to be very low.

Not saying you made the wrong decision, just that 20 data points is NOT enough to make a determination with any statistical significance.

One reason why a lot of campaigns fail is because they get taken down before they have a chance to succeed.

I blogged on the math behind testing yesterday. If anyone can tell me how to post a spreadsheet to a blog, I will post one that I know works without the need for complicated math or formulas.
 
Heh, weird... I do a lot of math to. You ever look at the standard deviation of your campaigns?
 
Josh P said:
yes, that offer did just suck.
and Sonic, 45 clicks is definitely enough statistical data. More than 20 data points is accepted in the world of statistics. And yea, I was on to something, I took what worked of the campaign, and changed some things, and now I have a working campaign.
Click to expand...

You can do what you like obviously, but 45 is not significant for PPC testing. What if that 46th click converted, then you have a CR rate of 2.1%, or the next 2 converted, you have 4.2%. I would say something on the order of 100-200 minimum clicks is required to get real "feel" for the market.
 
Josh P said:
yes, that offer did just suck.
and Sonic, 45 clicks is definitely enough statistical data. More than 20 data points is accepted in the world of statistics. And yea, I was on to something, I took what worked of the campaign, and changed some things, and now I have a working campaign.
Click to expand...

One should always run a ppc campaign for atleast a full 100 clicks. Like others have posted, you never know if the first 20-30 weren't interested to begin with. The full 100 will give you a solid conversion rate ( somewhat ). If after the first 100 clicks, your not converting change your shit up, tweak your ad, maybe make changes to your landing pages. Then try again. Dont work again, tweak some more. After 300 clicks and your still not showing a profit, scrap it and move onto another offer.

Josh, I see Dnic24 is helping you out pretty good. Nice to see you getting rolling! :drinkup:
 
Diorex said:
30 data points is typically the minimum required in statistics, and that is when comparing multiple test to determine a winner and even then your confidence level is going to be very low.

Not saying you made the wrong decision, just that 20 data points is NOT enough to make a determination with any statistical significance.

One reason why a lot of campaigns fail is because they get taken down before they have a chance to succeed.

I blogged on the math behind testing yesterday. If anyone can tell me how to post a spreadsheet to a blog, I will post one that I know works without the need for complicated math or formulas.
Click to expand...

Not always true.
 
Adsense Loser said:
Not always true.
Click to expand...

I agree that there are certain exceptions where less than 20 data points can be statistically significant (most notably a closed set), I disagree that when you are talking about internet marketing and a potentially open ended volume of data that 20 or even 50 occurrences is unlikely to show you an accurate reliable picture of the long term prospects on a consistent basis.

There are flaws and limitations to everything statistics related. This is not the forum to discuss them. The formulas I provided will give people a statistical basis for making decisions rather than a gut feel. Statistics is still going to be wrong some of the time.
 
Josh P said:
45 clicks, 5.03%CTR (this is very good)

Once they are one my site, which only has links internally (like 2) and links to the CPA offer.

Close to 50% of my visitors are clicking the link to the CPA offer, but of those, only about 9% are actually doing the CPA... so I'm looking at 4-5% overall conversion rate = mass suckiness.
Click to expand...

Couple of points...

5.03% CTR is kinda average, 80% is what I would call good :)

50% of your visitors clicking a link is also about average, so don't worry there.

9% of those completing an action is average to good.

So what I'm saying is welcome to the world of PPC to CPA. Your figures are normal, all you have to do is find a niche that pays well and converts. That is the hard part. And 45 clicks is nowhere near enough to gauge a campaign, give it at least 500, preferrably 1000. And don't forget about those cookie sales which could occur days or weeks after the clicks.
 
Status
Not open for further replies.
Top Bottom