[MichaelColey]

Stephen,

I just wanted to make sure you didn't miss my challenge here:

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by MichaelColey:
So you're saying that if I set my browser to not accept cookies at all, if I click through on a Linkshare merchant link and place an order, it will track properly? I would be very surprised if that's the case.

If you claim that it is, would one of my Linkshare merchants PM or email me about doing a test order?

The ONLY merchant I know of that works totally without cookies is Amazon. They embed a session ID in all their links. Perhaps some Linkshare merchants do that as well, but I would be quite surprised. Most merchants rely on cookies for session tracking. <HR></BLOCKQUOTE>Let me know if this is what you're claiming.

[smesser]

Jeff,

Thank you for clarifying that you agreed with me technically that if a user has cookies shut off then they will not be tracked. As I mentioned earlier we can deal with the other issues that relate around tracking after we get the basic facts. I am trying to make this easy for readers (technical and non technical to follow) and thus a step by step build up is easier. It would also be easier if everyone saved their arguments for why they think a cookie based tracking system is not completely flawed till the end.

Ok so now we have agreed that on an initial click, if a user has cookies shut off then no sale can be tracked. Let’s call this for our discussion, Cookie failure #1.

Now let’s discuss the next issue in the logic tree. On the initial click if a user has their cookie on. Are they still prone to having that cookie altered or deleted prior to sale?

For example, you mentioned Norton, but adware/spyware removal programs. Don't they delete cookies as well, but only third parties like other affiliate providers. So in other words, technically, wouldn't a user who has cookies on, still be able to buy but not be tracked by a cookie based tracking in this scenario, and wouldn't LinkShare patented non cookie approach be unaffected?

Again, let’s stay focused on the logic tree before we move into the spin for why anyone feels the loss is ok. I think that the affiliates here deserve a clear explanation.

Best regards,
Stephen

[smesser]

Michael,

I have not forgotten your challenge, I just want to stay focused on the logic tree discussion and then, once we are complete circle back to all the secondary questions. It is always so confusing for me (and probably others) to follow six threads at the same time. I just want to keep this simple at this time if possible. I hope you understand my hope.

Steve

[ToddCrawford]

How would commissions get tracked by LinkShare if cookies were turned off and the merchant themselves cannot process transactions without cookies turned on?

The fact of the matter is (as Jeff also pointed out) very few people have cookies disabled and those that do are probably not buying online anyway.

Every tracking technology except for LinkShare uses cookies to track activity. These include website analytics, keyword tracking programs, ad servers, affiliate programs, media networks, etc.

If the entire online advertising/marketing/e-commerce industry uses cookies why is LinkShare the only one that decided back in 1996 to develop technology that does not use cookies?

[MichaelColey]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Jeff_Kowabunga:
Let's assume 1% have cookies off (though it is lower in our research). Let's further assume that 80% of sales occur on the first visit. A quick comparison, then:

LinkShare 1,000 visits, 10 cookies off, 2 of those purchase after the first click and are not tracked by LinkShare.

CJ (assuming cookie tracking), 1,000 visits, 10 with cookies off who are not tracked by CJ.
Overall, LinkShare tracks 8 more visits out of 1,000 than CJ - 0.8%. <HR></BLOCKQUOTE>Actually, when you factor in the 71% of Linkshare merchants whose shopping carts won't work without cookies, it drops the number down to about 0.2%, FAR below the 10% that Stephen claims.

[Jeff_Converseon]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Thank you for finally admitting that if a user has cookies shut off then they will not be tracked. <HR></BLOCKQUOTE>
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Again, let’s stay focused on the logic tree before we move into the spin for why anyone feels the loss is ok. <HR></BLOCKQUOTE>

I don't think my post was spin at all.

You asked me to answer your original question, which included an assumption that 10% of users have cookies off. That assumption is incorrect, and I explained how that assumption would cause a misleading conclusion.

I don't think anyone here ever believed that users who have cookies off could be tracked by an application that needs cookies; that is basic logic.

To answer your next question, I am sure that there may be some internet applications that "help" users by automatically setting their IE security settings to high. Again, this only blocks third party cookies without a proper privacy header in place, and we all have them. There may also be applications that just block all third party cookies altogether. As I said before, a smart merchant or affiliate would have to balance this risk with other risks inherent in other tracking methods.

I do agree that we need to stop spinning this. It's really a very simple issue. I do think MichaelColey has a good question, however. If LinkShare's "non-cookie" tracking requires the merchant to pass that id from initial click to sale event, and that merchant doesn't pass it through the URL (like Amazon), then are they using a first party cookie/session to do this? If they are, then LinkShare's system is just as reliant on cookies as anyone else's.

Not trying to be confrontational here (I think this thread has had enough of that); but since you are so adamantly opposed to any provider using cookies, I think that is a crucial question.

[smesser]

Todd,

My example above covers the second scenario in the logic tree. That second part of a logic tree covers people who have cookies on, but those cookies are altered by third parties (of which anyone on these boards can cite numerous such programs in existence).

In this scenario, a user can shop without problems, but they are no longer tracked by third party cookie technologies. So taking this scenario, I am simply pointing out that even when cookies are on; they are subject to many significant tracking errors. Errors that our users are not subject to because our tracking does not rely on cookies.

Todd, with all due respect, I have seen you mention now more then once that all other tracking providers use cookies. For one thing, this would fact would still not under cut my argument about inaccuracy in tracking. But secondly I would argue that almost all of those people do not compensate on sales. Most use it to track CPM, CPC deals where the value is always the same. In other words a failure counting the number of impressions is a consistent and numerically manageable issue. In dollar transactions, a missed sale could have been for $10 or $100 or $1000. A missed sale has very different monetary implications and thus I would not say that this discussion would add much weight to your arguments.

Again, I am simply trying to follow a clear logic tree and save the arguments for the end.

Best regards,
Stephen

[ToddCrawford]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>But secondly I would argue that almost all of those people do not compensate on sales. Most use it to track CPM, CPC deals where the value is always the same. <HR></BLOCKQUOTE>All use it to track the event/transaction, which is important for determining ROI. It doesn't matter if their business model pays on CPM, CPC or CPA, they are tracking via cookies.

On more point to clarify (again), CJ sets cookies a a first party, not a third party.

[smesser]

Jeff,

Just to clarify, my assumption was not meant to state any of those numbers as fact. I simply used round numbers to illustrate simple (and easy math) around my point. That is why all the numbers where built around 10 of some kind. While I would argue, and most third party companies show higher percentages of users (then you or Todd claim) who shut off cookies, I was saving that argument for the most appropriate place on the logic tree.

Just so everyone knows; I am following a logic tree which is a traditional software developmental process to identify flaws in programming logic or business logic. This is a simple way to follow every issue to its logical conclusion. It also helps to keep the discussion clean.

Best regards,
Stephen

[smesser]

Todd,

In this case I am reffering to software applications recognition of all of us as third parties. I did not mean third party cookies in the sense of IE. I will be more careful with my choice of language in future posts.

Steve

[smesser]

Hi all,

I have to step out for a few hours so can I ask that we put this discussion on hold until I can join back in. I apologize again, but as I mentioned I had scheduled other meetings and was not expecting to do this today.

Best regards,
Stephen

[smacie]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by smesser:
Just so everyone knows; I am following a logic tree which is a traditional software developmental process to identify flaws in programming logic or business logic. This is a simple way to follow every issue to its logical conclusion. It also helps to keep the discussion clean.
<HR></BLOCKQUOTE>

Stephen,

As someone who actually is a software developer I thought I would add a clarification.

It is true that developers follow a logical progression when solving problems but we also know enough to examine which are the big problems and which are the little ones. We would never start by looking at a scenario which happens 0.2% (thanks MichaelColey) of the time. You start by looking at the big picture.

Let's be honest here. Your argument is not a logical one. It is a political one.

[ecomcity]

For all of us digesting this cookie vs proprietary tracking information interchange could the networks clarify the level playing field. Do all tracked affiliates sales, both from domain bound and roving S/W BHO affiliates, use the exact same network sales crediting system???

[smesser]

Hi all,

I apologize for not getting back sooner, but with everyones permission can we continue this on tuesday. I am just a bit behind today. I will try to block more time on tuesday to finish up this topic.

Best regards,
Stephen

[SSanf]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> Originally posted by ToddCrawford, CJ.com:
Either way, sessions can be impacted by opening a link on the site in a new window. <HR></BLOCKQUOTE>

So, what you are saying is that if a merchant has the cookie set at keep=no and the customer opens a new window in order to not lose the page for an item they might want, there is no chance of the affiliate being paid, correct?

And, since most people do open a new window at some point in their shopping, keep=no merchants get most of their sales by passing CJ and the affiliate, isn't that right?

That probably accounts for a whole lot of the sales people say they made that weren't tracked!

CJ should NEVER allow keep=no!!!!

[ToddCrawford]

Ssanf,

There is no such thing as keep=no anymore. All cookies are kept. In each program term, advertisers determine whether they pay on one, all or some number (&gt;1) of txns during their referral period.

Opening a new window does not effect tracking via cookies.

I am talking about session-based tracking, where the advertiser either keeps track of the session with a cookie and/or captures data and stores it on their servers to track transactions. Depending on how they keep track of the session, it could get broken if someone opens a link in a new window, closes the browser or leaves the website.

[Jeff_Converseon]

I am fairly sure that a session being lost when a new window is opened was only a problem in early versions of IE 5. IE 5 Sevice Pack 1 had a fix, and it has been fixed since then in later versions (which almost everyone uses). http://abw.infopop.cc/infopop/emoticons/icon_smile.gif

http://support.microsoft.com/default.aspx?scid=kb;en-us;300895

[SSanf]

Well, I am not sure I understand what you are saying. Is it possible to not get paid if someone opens a new window?

[Kellie aka Ms. B]

Here's my question for Todd and Steve for what is the most glaring question to be answered based on what I'm seeing so far regarding the differences in tracking methods between CJ and LS. Sorry if I'm skipping a few branches in the logic tree Steve. http://abw.infopop.cc/infopop/emoticons/icon_wink.gif In an attempt to try and keep things siimple, it's a multiple choice question. http://abw.infopop.cc/infopop/emoticons/icon_smile.gif

Q: In any situation where the Network cookie is either blocked from setting or deleted from the end users computer prior to the end user making the purchase, will your systems track the sale? Any cookies set by the Merchant are remaining intact. The time of purchase is anytime from a second after the very first click from an affiliate link to the last second of whatever the Merchant's return days are. We are assuming here that Network link was there to click on (blocked links of course impact both Networks equally since there is nothing to click on) and the click went through the respective Network's servers. We are also assuming that the Merchant has implemented all tracking on their end per each Network's specifications (since that is another issue).

A: a) Yes it would track.
b) No it would not track.
c) It would track in some situations. If c) please explain briefly the circumstances in which it would track.

[Jeff_Converseon]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Q: In any situation where the Network cookie is either blocked from setting or deleted from the end users computer prior to the end user making the purchase... <HR></BLOCKQUOTE>
In order to keep us to the logic tree, I'll help with that one http://abw.infopop.cc/infopop/emoticons/icon_smile.gif

The question doesn't really apply, as LinkShare doesn't set a "Network cookie". I think Steve has been pretty adamant about that, as he is trying to build a case that cookies are an inferior way to track affiliate links.

The more important question is whether a user who has cookies off can be tracked by LinkShare, as it would seem that the LinkShare merchant probably needs to set their own session cookie in order to get the LinkShare id from click to sale event. I'm assuming that's the question Steve will be addressing today.

[Kellie aka Ms. B]

Jeff, I understand what is being said by Stephen. I worded and asked my question exactly how I intended to and feel it does apply for what I am asking and wanting to know. http://abw.infopop.cc/infopop/emoticons/icon_wink.gif

[smesser]

Good morning everyone:

I apologize again for having to run in and out of this discussion. I will do my best today to stay in it as long as I can but I hope everyone knows that this may not always be possible as I have other meetings during the day. I will let everyone know when I have to go, like yesterday if that worked ok. Also I appreciate all of you giving me so much of your time and being patient with my approach to clarifying the issues.

With regards to Ms. B, if I understand your questions correctly, that being a user who clicks on a tracking link, the network’s cookie is not set (for whatever reason), but the merchant’s cookie is set, would LinkShare still be able to capture the sale. By the way, I disagree with Jeff’s response to Ms. B, as she is dead on target with her question based on the logic tree we have discussed. If that is the question, then technically, the answer would be yes (letter A in the multiple choice selection), LinkShare would capture that sale

With regards to Lurking programmer, I see that you are new with this being only your third post so welcome. I would disagree with not focusing on the largest risks. The initial click (whether with or without cookies on) is the bulk of all of our traffic, and in fact is the basic requirement for all of us to credit a sale to an affiliate. Your argument seems to focus on the numbers of people with cookies shut off and Todd’s unscientific test of merchants using cookies for shopping multiplied in some format. Since we have not gotten to arguing about scope of problems yet I will reserve judgment on that. I think when we are done you will see what I mean by this.

So far we have identified and agreed on two known tracking flaws in cookie based tracking.

Cookie Flaw #1 is simply when a user has cookies shut off
Cookie Flaw #2 is when a user has installed adware removal or some anti virus programs such as Norton that deletes all cookies set by third party advertising companies. (To be clear I am not referring to third party cookies in the IE sense of the term)

As I mentioned the prior two issues do not affect LinkShare’s tracking because we do not use cookies to track sales. If anyone thinks that there is a technical error here please let me know and we can discuss that issue further. If not lets move on to the next area.

So now we are on to Flaw #3 in cookie tracking. Since we have already discussed the risk of adware removal programs (Lavasoft, spybot etc.) deleting all but the merchants cookie, let me stay with this theme for a moment longer.

To collect the data at the end of any sale, cookie trackers rely on a 1x1 pixel to retrieve the cookie and thus complete the collection of sale data. A basic cookie rule is “only he who set the cookie can retrieve it”. The 1x1 pixel allows the cookie tracker to retrieve their cookie and thus associate the sale to the right partner. If for whatever reason, and these reasons happen more often then you would imagine (network congestion, server crashing, user error, Norton blocking the link or a merchant who removes the link or worse yet randomizes the page with the pixel so that sales are randomly reduced) the sale information will be lost forever. There is no way to retrieve this information and in this instance tracking is gone forever.

Since LinkShare uses a merchant set cookie for return days, and since our data is stored in the merchants database with shopper/cart id associated with it, we create a audit log that does not rely on a cookie (again, return days are the exception). Even with return days we do not need to have a 1x1 to collect that information, the merchant server, which has to be up to complete the sale, does that for us.

So Cookie error #3 would be the pixel not working (timing out, or some other related error discussed above). In this case, once the user has made a purchase, that sale is lost forever to the cookie trackers. There is no way to go back and fix this problem.

Please let me know if you agree with this statement.

Best regards,
Stephen

[Kellie aka Ms. B]

Stephen, yes that is exactly what I meant. But, why the "technically" part to your response (so everyone reading the thread is clear)? With regards to "cookie flaw #2" as you are calling it and since LS does set linksynergy cookies (granted you are saying that LS's tracking is not dependent on such cookies), is there anything in LS's reporting which gives an indication to the frequency (as in a %) that "cookie flaw #2" actually occurs?

[Jeff_Converseon]

First, regarding Ms. B, I do apologize for any misunderstanding; her question asked about LinkShare's network cookie, and LinkShare doesn't actually use one -- just trying to make sure we all understand the issues.

And Steve is absolutely correct in the following "cookie flaws" regarding third party cookie tracking:

1. If a user has cookies off, they user will not be tracked.
2. If a user has Norton installed, and Norton is set to block third party cookies (though I am not sure that this is default), then the user will not be tracked.
3. If the pixel tracking tag does not load, the user will not be tracked.

And a quick quote from Steve before I continue:
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>As I mentioned the prior two issues do not affect LinkShare's tracking because we do not use cookies to track sales. <HR></BLOCKQUOTE>

I was under the impression that this thread was started in an effort to understand how LinkShare uses cookies to track sales. LinkShare's position has been that cookies are an inadequate way to track, and tout their system as "better" because it does not use cookies.

Yet it is my understanding that a LinkShare merchant must set a session cookie in order for the LinkShare sale to be tracked. It is the way that the LinkShare click information makes its way to the sale event and the final cart contents. Without that cookie, the LinkShare sale will not track. In other words, LinkShare, though it does not set a "network" or "third party" cookie, certainly does use cookies to track sales.

If I am correct about that (and please let me know if I am not), let's take a look at the 3 cookie flaws Steve has confirmed, ad how they affect LinkShare:

1. User has cookies off: the merchant session cookie will not be written and the LinkShare sale will not be tracked.
2. User has Norton Internet Security: the LinkShare affiliate link will be blocked, and thus could never be clicked on in the first place.
3. The image tag doesn't load: As Steve has pointed out, LinkShare does not use an image tag to track sales, so this would not apply. However, LinkShare does require the merchant's cart and session cookie to work correctly in getting the affiliate information to LinkShare. This process not working correctly could be as likely as an image tag not loading, though I have no data on that.

All of that being said, I think LinkShare's method of tracking is excellent. As I've mentioned previously in this thread, we've had merchants request a simliar tracking from us for their programs, and we have found it to be very reliable. However, I think it is misleading to say that it does not use cookies.

[smesser]

Ms. B,

Nothing more was meant by the term technically, other then I was trying to be exact in answering your question. With regards to frequency, I am going to handle that issue shortly so I will share more information then if that is ok. I just do not want to get ahead of myself.

Best regards,
Stephen