ReveNews
-
NAVIGATION
-
TOPICS
-
THE REVENEWS BLOGGERS
-
QUICK CONTACT
LinkShares Sales Reporting Web Service and Affiliate Tracking Methods 101
Shawn Collins from AffiliateTip.com reported about the new LinkShare web services tool for sales reporting by Advertisers. After reading the post did it occur to me that many (including Shawn how it seems) do not understand how the affiliate tracking works and what the difference is between the two main methods that are used by affiliate networks.
Affiliate Tracking Methods
Commission Junction offers advertisers the choice between pixel tracking and batch processing for the affiliate tracking implementation. Pretty much every network except from LinkShare offers pixel tracking. Some networks even offer pixel tracking as the only available tracking option, especially if they target small and mid-size advertisers only. LinkShare on the other hand does not offer the pixel tracking option. This is one of the reasons why the minimum initial implementation cost of the LinkShare solution is more expensive than the initial cost of the pixel tracking implementation by any other network. Batch processing is more interesting for large advertisers, but it is also more complicated and harder (and more expensive) to implement for a merchant.
The Pixel Tracking Issue
Tracking cookies created by networks on computers of users, which are needed for the pixel tracking method became threatened by spyware and anti-virus software vendors who started to flag those cookies as unsafe and then deleted them from the user’s machines. Click2Customers hired Ben Edelman to do some research of how bad the situation actually is. The result of this research was a special report that Clicks2Customers released to the public for free in September 2006.
How Does LinkShare Track?
When the issue of deleted affiliate cookies was debated did LinkShare always claim that they are not affected by this, because their “cookie-less” tracking system did not depend on this. Everybody was speculating how else they do the tracking, especially for programs that offer 30 or more return days. Finally in 2006 did LinkShare’s CEO Steve Denton came out of the closet and explained in an interview with Wayne Porter how their tracking-system works. They also use cookies, just not on the network side and there is no magic that was only known to LinkShare to make their tracking work. We learned that LinkShare does also just “cooks with water” like the rest of us. 
Pixel tracking? Batch Tracking? What does this actually mean and what are the pros and cons of each of the methods? I think it is about time for some clarification for the non-technical folks.
Pixel Tracking
- User clicks on affiliate Link that refers to a domain owned by the network. Prominent tracking domains include Qksrv.net (CJ*1), click.linksynergy.com (LinkShare), service.bfast.com (BeFree), ww.shareasale.com (ShareASale) and clickserve.cc-dt.com (Performics)
- The network records the click, sets a tracking cookie on the users machine and redirects to the specified target URL at the advertisers website
- User lands on the advertisers website
- User shops at advertisers website
- User completes a purchase and hits the confirmation screen that thanks the user for the order and provides him with the order number. This page contains an invisible 1×1 pixel image*2 that is located on the networks server and is not loaded from the advertiser’s website. This image is actually not an image, but a tracking script that records the transaction.
Notes:
*1 Commission Junction owns 86 domains, where most of them are used for tracking
*2 1×1 Pixel tracking example codes
Commission Junction 1×1 pixel tracking code example
<img src="https://www.commission-junction.com/track/receive.dll?AMOUNT=100.00&CID=999999&OID=100001&TYPE=1111&KEEP=YES&METHOD=IMG" height="1" width="1">
ShareASale 1×1 pixel tracking code example
<img src="https://shareasale.com/sale.cfm?amount=1.00&tracking=1000001&transtype=LEAD&persale=&perlead=&perhit=&merchantID=99999" width=1 height=1>
Batch Tracking
- User clicks on affiliate Link that either refers to a domain owned by the network OR directly to the advertisers website
- (optional) The network records the click and redirects the user to the specified target URL at the advertisers website
- User lands on advertisers website who sets a cookie at the users machine for tracking purposes
- User shops at advertisers site
- User completes a purchase and hits the confirmation page at the advertiser’s site. The advertiser saves with the order relevant data for the affiliate program tracking.
- (offline) At this point is the user not on the advertiser’s site anymore. The advertiser sends recorded transactions that involve an affiliate to the affiliate network.*1
Notes:
*1 The transfer of data is usually done via FTP transfer on a frequent basis. The frequency of the transfer varies from advertiser to advertiser. Some do it every other day, maybe even only once or twice per week, while others do it daily or multiple times per day. It is never done in real time for a single transaction.
Pixel Tracking Pros
- Easy to implement for the advertiser, because it only requires a somewhat minor modification of one page of their website to make it work. Everything else is setup and configured on the network side, by the network itself or the advertiser via the backend interface.
- Instant recording of a sale, which is often shown instantly (or only delayed a little bit) in the reports for the publishers/affiliates as well.
- Advertisers can’t manipulate manually recorded sales. They could do something programmatically, but that would raise suspicion fairly quickly.
Pixel Tracking Cons
- Vulnerable to fraud and hacking
- Dependent on network cookie, which is blocked by a number of anti-spyware and anti-virus tools or browser plug-ins.
- Advertiser has to report sales that are not fulfilled yet and might get cancelled, because he cannot delay the transfer of the transaction to the network. This increases the likelihood that he has to reverse transactions manually via the admin interface of the network visibly to the affiliate.
Batch Tracking Pros
- Advertiser has more control of what transactions are sent at what time to the network. This allows the advertiser to withhold transactions until the sale is actually completed and reduces the need and likelihood of reversals of transactions that were already reported to the network and seen by the affiliate.
- The tracking cookie to allow return days beyond the current browser session (also called cookie-duration) is set on the advertiser’s domain. It is less likely to be blocked by any third party software or being deleted manually by the user because the user does more likely recognize the advertiser’s domain than the tracking domain of a network.
Batch Tracking Cons
- More changes and additions to the advertiser’s website code and business processes are necessary, which require deep technical knowledge and higher sophistication for the proper implementation of the tracking mechanism
- The advertiser has control over what is being sent to the affiliate network and when it is being sent. Fraudulent activities or simply bugs in the tracking code on the advertiser’s website are harder to detect.
- Reporting data are not available to affiliates quickly. This can become more of an issue the less frequent an advertiser sends transactions via batch to the network. Imagine a paid search affiliate who runs a major campaign for an advertiser and has to wait a day or longer to see if the campaign actually works or if money is wasted.
What Does the new LinkShare Web Service?
So what does the new LinkShare web service do and change? It addresses the “con #3″ of batch tracking.
The web services allows instant reporting of transactions to the network, eliminating the need to wait for the next scheduled batch transfer. I assume that the web service is optional for LinkShare advertisers and can be used to replace the batch transfers entirely or can be used in conjunction with it to send some transactions immediately while still sending other transactions via batch transfer later, if needed. This is a good feature, because it offers advertisers a choice and the ability to take only the good things from each method without getting the bad things than usually came with it, except for the higher sophistication and complexity of the implementation of course.
Death of the Pixel?
Shawn asks the question
“Does this mark the death of the pixel?”
The answer to that question is “No“.
Small advertisers will still go for the cheaper and easier to implement pixel tracking method. If CJ would offer a service like this, it could become another good reason for larger CJ advertisers to switch to batch processing, advertisers who hesitated doing it because of their paid search affiliates. The only thing this announcement by LinkShare changes is that we will see more advertisers on the LinkShare network reporting sales data in pseudo real-time back to the network who is then able to show those data also in pseudo real-time to the affiliate/publisher. Search affiliates and others will welcome this change and appreciate it.
Further Reading
I have CJ’s batch processing implementation guide (advertiser action data feed) and LinkShare’s integration guide available on my site, together with other guides for data feed and web services implementation for various networks.
I hope that I can get my hands on the web services guide for advertisers to report sales data back to the network. If you are an advertiser who plans to look into this and requested the integration document from LinkShare, please contact me and send me that document. If you are working for LinkShare and willing to send me the document that would be even better. I would appreciate that. Thank you.
Carsten Cumbrowski
> After reading the post did it occur to me that many (including Shawn how it seems) do not understand how the affiliate tracking works and what the difference is between the two main methods that are used by affiliate networks.
Carsten – the information from my post was all taken from the release from LinkShare, so I’m not sure how you surmise that I don’t understand the nuances of tracking.
And when I ended with “Does this mark the death of the pixel?”, it was meant to be rhetorical to echo the desire of many affiliate marketers to evolve to a more consistent tracking method.
Hi Shawn, thanks for the clarification.
I suspected this actually, but wasn’t entirely sure. So I thought that I will mention it and see if and what you will respond. I just wanted to keep you on your toes
Btw. What do you think about my general explanation of the tracking methods?
Cheers!
Carsten
I will just say Linkshare is in the dark ages when it comes to reporting. SAS and Performics, pretty instant. CJ usually within 2 hours unless it’s a batch merchant. Linkshare, whenever the merchant sends in the data. Some send it in more frequently that others. But affiliates in general like as real time as possible reporting.
Nice explanation – perhaps a bit elaborate for a newbie trying to wrap their head around it.
Thanks for the honest feedback Shawn. I left a bunch of stuff out actually (no kidding). You could consider this post the “short version”
. I know, I know. I hope that my colorful illustrations will help with the understanding of those things. They are technical in nature, which is always tough to explain to a non-technical person.
Another “con” with batch tracking is that every landing page on the merchant’s site has to be set up to track affiliate traffic (most often by setting a cookie that identifies the affiliate, or logging the affiliate data into the session). I’ve run across numerous situations where batch-tracking merchants added a new links/page which didn’t properly track. To me, this is one of the biggest problems with batch tracking.
I’ve long been an advocate of redundant tracking (using both a network tracking pixel and merchant-side tracking). Although I’ve proposed this to virtually every network I work with, I don’t know of any who have implemented it (or even have it planned for the near future). I did have one merchant who had both forms of tracking in place for a few months (while transitioning from one to another) and I found that both tracking methods caught about the same percent of sales that the other had missed. I don’t remember the exact details, but it seems like either method by itself missed almost 5% of sales. It’s far from being a scientific study, but I think it’s fairly representative.
Hi Michael,
This is actually included in Con 1 for batch tracking, but I did not spell it out the way you did. Thanks for clarifying that for people who are not aware of that.
I hear you regarding the redundant tracking. I did that for BevMo.com actually when I managed their program. They decided for CJ pixel tracking and that was the basis for the affiliate commission. We did however enabled in CJ that the AID and PID is transmitted to the destination page at BevMo.com and set our own tracking cookie and recorded that with sales and new customer sign ups.
To answer your question you probably have. No, the numbers did not match up. But that came not to a surprise. It was almost always the case that CJ was not recording the sale instead of the other way around if there was a difference. This was most likely caused by a deleted CJ cookie. The difference over all was not bad enough to raise major concerns back then. I would have loved to dig into much greater detail, but time did not permit me to do that.
What I got out of the general numbers was that the cookie blocking issue was not as big of a deal as some people wanted to make others believe.
I would recommend to anybody to do your own tracking next to pixel tracking. We discovered fraud that way. I will not post here how the fraud was done, because I learned that this works on other networks as well that use pixel tracking. I tested it myself at ShareASale where I have affiliate accounts and an old advertiser account. I sent Brian an email back then (2003 or 2004 I think), but he never responded to my detailed email that contained what I did, what the results were and some suggestions to detect at least some of the fraud automatcally.
. I could test SAS though. Maybe I will. mmh.
I have not tested in lately, but I would not be surprised if it would still work the same way. I also ran out of a CJ advertiser account to conduct a test
Nice article. I use both methods for our product sales with batch tracking used as the official numbers . Pixel tracking has shown to be consistenly 5% less than our internal tracking. For the most part,I think the lack of accuracy is just part of doing business online much like click fraud with PPC. The good news is the inaccuracies are fairly consistent so it can be mitigated.
Hi Kevin,
thanks for your comment and for sharing some of your numbers.
Based on the calculator that was created by Ben Edelman as a result of his study about cookie removals, does your 5% difference between your internal tracking and the networks pixel tracking indicate that 80-85% of conversions occur on the day of the referral. I know that ShareAsale provides a report that shows the percentage of conversions during the cookie duration. Other networks do not, but you might do this tracking as part of your internal tracking.
Can you confirm this 80-85% estimate for conversions that happen within the first 24 hours after the referral based on your own experiences? I appreciate it.
Its too bad that not many advertisers do this kind of double tracking and share some of the general figures.