What Ratio Do You Value?

[jayeff]

Conversion ratios are almost totally meaningless:

1. Because of differences in the way sponsors record their stats, at the very least you would need to compare sales against click stats of your own.
2. Why should conversion ratios be consistent when promotions run from freebies to high-cost products and cover everything between?
3. What do conversion ratios tell you unless you relate them to income: particularly with recurring programs?
4. If you want to see just how meaningless conversion ratios are, stick up a blind link in a prominent position. Your conversion ratios will go down the toilet, but your income may actually increase.

People focus on conversion ratios, no matter how many times they are said to be near useless, because they can. And hey, better to be doing something and better still if it's easy. The only relationship which really matters is earnings per impression, because two things determine your income: traffic and the effectiveness of the promo spots you display to that traffic.

Unfortunately, it isn't a simple calculation. You should reasonably expect prime promo spots to perform better than secondary ones, big promos to do better than small. Thus you cannot say I showed Sponsor A's promo the same number of times as Sponsor B's, but earned more from B, therefore he is better. You need to either figure out a balancing factor or compare the results of sponsors to whom you have given a similar "quality" of exposure (ie one group for those to whom you gave a space of their own at the top of your page, another for those appearing only in a list of links, etc).

Not straightforward, but then why track stats at all, unless they actually mean something?

[Licker4U]

OK, so based on my numbers above, which is for one week, that's .01/click. I'll check some monthly and six month totals before I move on. I know you can go two weeks with nothing then boom, several sales back to back.

[Greenguy]

Originally Posted by Licker4U

OK, so based on my numbers above, which is for one week, that's .01/click. I'll check some monthly and six month totals before I move on. I know you can go two weeks with nothing then boom, several sales back to back.

If you're 1:3502 now, I would not go much past 5000 hits before I made my judgment (if I didn't pull them today, because that ratio is horrible)

[Licker4U]

Originally Posted by Greenie

If you're 1:3502 now, I would not go much past 5000 hits before I made my judgment (if I didn't pull them today, because that ratio is horrible)

I know, that ratio is worse than a link list ratio

[Licker4U]

I need to do this more often. Since Oct. 21, 2006 (sponsor stats don't go back any farther)

1st page hits: 51,538
2nd page hits: 6,363
Join page hits: 11,974
Sales: 22 |shocking|

Without a calculator that's .012/click. I should have done this months ago

I saw small checks dribbling in and thought things were OK.

(note to self: check $'s/click, whatever clicks the sponsor shows)

Where's the smiley kicking himself in the butt?

[Linkster]

Most important - and impossible to tell without a good clickout script on your server - is what was mentioned before - but not comparing them to the sponsors click tracking but to your own - then you really know what your earnings per click(impression) are.
Unfortunately most people still trust what the sponsors click tracking reports - and as seen recently this can be sadly way off base even with large companies like Google

[Greenguy]

Originally Posted by Licker4U

I need to do this more often. Since Oct. 21, 2006 (sponsor stats don't go back any farther)

1st page hits: 51,538
2nd page hits: 6,363
Join page hits: 11,974
Sales: 22 |shocking| ....

Before you dump the sponsor completely, I'd run stats using those time frames for the individual sites, as you might be converting good with some of the sites in the program, but not others.

I do that all the time - there are some sponsors where I only promote some (or even one) of their sites as they do convert well with my traffic while the others convert like shit.

[Licker4U]

Originally Posted by Linkster

Most important - and impossible to tell without a good clickout script on your server - is what was mentioned before - but not comparing them to the sponsors click tracking but to your own - then you really know what your earnings per click(impression) are.
Unfortunately most people still trust what the sponsors click tracking reports - and as seen recently this can be sadly way off base even with large companies like Google

LOL How could my stats get any worse? If they're not counting all the hits I'm really fucked.

Originally Posted by Greenie

Before you dump the sponsor completely, I'd run stats using those time frames for the individual sites, as you might be converting good with some of the sites in the program, but not others.

I do that all the time - there are some sponsors where I only promote some (or even one) of their sites as they do convert well with my traffic while the others convert like shit.

Now that's an idea. That's gonna take some time but since I'm not making any money what else do I have to do?