Affiliate program reporting inconsistencies
 

Recently I got a programmer to design me a redirect script that not only forwards a visitor to an affiliate program but it also logs referring urls, raw hits, unique ips, dates, times etc...The whole idea behind this is so that I can compare my statistics against the affiliate program I'm promoting.

Well to cut a long story short, I'm logging 10% or more unique IP's each day compared to the affiliate program stats.

I have exported the data and given it to my rep, but he has replied with a bullshit answer on why this is occuring.

He has promised that every click no matter how it is tracked that results in a signup guarantees that I will get paid for it.

I just don't feel so sure.

What do you think I should do?

Send to a different program.

For the record, as an owner of a program and as an affiliate - we all know variances in clicks DO occur. It's up to you to decide what threshold is acceptable to you.

Would you call this acceptable?

30th Jan
My stats - 2134 uniques
Their stats - 1619 uniques

31st Jan
My stats - 2321 uniques
Their stats - 1862 uniques

1st Feb
My stats - 2312 uniques
Their stats - 1885 uniques

2nd Feb
My stats - 2296 uniques
Their stats - 1849 uniques


Do you honestly think I'm being credited for all signups even though our stats are so far apart?

Sorry but this is amazing. 2002 with only 2 posts. I'm sorry this program is nicking traffic off you.

LOL

they call this a "glitch"

its second to the word affilioates fear the most, "shave"

It is a pretty close average distance between your stats and theirs. You positive that everything is being done right on your end?

if you use the same techniques to track your traffic sent to other sponsors, how are the results?

Good question. I'm going to test this out tomorrow. I'll let you know the results.

*thread bookmarked*

I'm more of a reader than a talker! :-)

10% discrempency is actually pretty good, I typically get 15-25%. Hell on adsense I get up to 40% difference from Google's reports compared to mine.

While it does suck to see the differences, there's not much you can do about it. The way I look at it, is compare how much revenue you get from them based on how many hits your system logs. Figure the revenue per click. Then go with whichever sponsor gives your the highest rate. I don't bother looking at what their hit count is at this point.
WG

Agreed, there aren't enough variables. Testing on some more sponsors will across different billing platforms will make it a little more clear if it's the stats system/sponsor, or if it could be your script. :thumbsup

Can't wait to hear how the rest of the testing goes.

it could be latency, so the surfers are closing the page out before it loads?

Are your scripts counting SE bots? Some program owners weed them out... etc...

Someone might say that bots don't pass the referrer... And I would say NO SHIT.. But they do pass the GET request with your affiliate code when they surf the link.

lol...


Should be interesting to see your results across different sponsors.

are they counting second page uniques or first page?
how often to they mark it as unique versus how often you do? What do you qualify as a unique? 24 hours? 48 hours? one week?
I would ask they what they call a unique versus what you are calling a unique

I define a unique as being a unique ip address per 24hr period. They have indicated that is how they define it as well.

Yes, my script counts SE bots. My rep hasn't indicated whether or not they weed out SE bots. I won't bring this up unless he does. If I did then that would give him an excuse.

I caught a few that manually erase sales... since then I now make screen caps so I can post it here when they do it again...

Under normal circumstances, like a chargeback or refund, it would show up in detailed stats, but that's not the case here.. it simply just disappeared because detailed stats shows NOTHING.. Naturally they deny it or claim it didn't happen

As far as your issue goes, I've seen this occuring on many programs stats and they pretty much never match up as I've conducted similar tests and never found one to be an exact match.

I've noticed the same kinds of issues when tracking clicks onthe server and comparing them to a sponsor.

I saw hits going to a program that I know doesn't shave, and there was a difference of around 10%

part of it could be the browser is not handling the redirect properly, spyware that is hijaaking, DNS servers not responding correctly.

The rep at UltraDNS was trying to sell me on the concept that using their system, traffic would increase... maybe its due to DNS on why some traffic isn't getting to the sites.

Fight the redirect to no where!

I agree with WiredGuy; 10% isn't bad, it's about average from what I've seen.

Maybe I should just concede that's the way it is and move on. It would be good if our stats matched, but it doesn't look like this will ever happen.

I would love to know how to fix the problem as well, if you ever come up with a solution, please let me know.
WG

Do you do a single redirect to the sponsor? or do you do a double redirect and only count the 2nd one?

There are a SHIT LOAD of bots out there that hide themselves using "real" user agents etc. and they won't neccessarily following a redirect and thus the discrepency. 10% is pretty typical.

If you want to run some tests, put in a second redirect and only count the click after the first one. You can clean it up even more by making the first redirect a javascript one and also set a cookie. That way, you can scrub the click before sending it onto the program. That should tighten up that percentage and give you a far more accurate count of "real" surfers.

As mentioned already, any visitor that passes through two tracking type scripts has always shown a loss for us of anything around 10%. For example, visitor arrives from a website via a click, he hits the geo script and then quickly is directed to the chosen destination.

The whole process takes just a fraction of a second but the loss for us is often greater than 10%.