Storing a Cookie on PPC Traffic to Your Affiliate Site

[maxor]

Say you've got an affiliate site and you're sending PPC traffic to said site. Is anyone writing a cookie to the user's that are coming from the PCC campaign?

I see two potential uses for this.

1. If you write a cookie to the user and store a corresponding key in your database you could easily store information about this user and call it up if they were to return to your site.

Say for instance that I am a PPC traffic visitor and I'm viewing your site and I bookmark it and plan to return later. Upon my return, knowing that I've been here before, you could display copy that is more along the lines of "Welcome Back" type stuff. Additionally you could make use of some type of "This offer expires soon!" copy to encourage them to take action now.

2. More advanced than the 1st option, if you have a visitor viewing content on your affiliate site and you make note of which particular pieces this visitor is viewing the most, you could display that content in which they seemed interested more prominently and perhaps also with the "offer expires soon" copy upon their return to your site.

This all hinges on a few things:

A. Would this violate the TOS for your affliate program or your PPC account? (probably not)
B. Is it likely that a visitor would actually return to your site if they did not take action the first time?
C. Would more personalized, "expires soon" or suggested content based on a visitors past visits have any affect on converting them?

 

[chatmasta]

You may as well do it. But there's no reason to unless it's the type of site that people WILL return to.

So if you were sending PPC to something like a reference site that they might come back to in order to check stuff, yeah then do it.

But if it's just a landing page for some ringtones that they will either buy from or never come back to, don't bother.

 

[maxor]

Good point.

Another WF member posted some screen shots of their credit card offer affiliate site which seems like the type of place that somebody might bookmark and return to at a later date. That's what sparked this idea in the first place.

 

[on-on]

Another thing to consider, and your mind is working in the right directioin, is the value of this if you run a network of sites and can figure out some clever ways around the old domain/cookie barrier. There are all kinds of goofy postback, background posting and redirection hacks to this. I've known at least one whiz kid who claims to be able to use an AJAX solution to access cookie data across domains, but I'm dubious until I can sit down and prove it (and we're not to that point yet). Third party cookies are almost universally blocked/blockable (at least since Safari, FF and IE6) these days - certainly enough that you can't rely on them - so when someone claims a whiz bang cookie solution I don't hold my breath.

Anyway, marrying a central database with a network centric approach to running your various sites gives you a lot of flexibility that I'm sure one can imagine. You know, what if all of a sudden your sites automatically knew who someone was, what terms they had searched to find any of your sites, what offer verticals they had expressed interest in, etc. Some heuristic logic governing the way sites and pages work suddenly allows for self-aware sites that re-arrange their content to support the expected user affinities. Of course it's a ton of work and it's not magic - and Amazon.com has more than once decided that I am a gay Nazi (last time I browse for "Keep the River on Your Right" and History Channel documentaries back to back) - but anytime you can match demonstrated user preference to targeted content over a large enough number of transactions to iron out the statistical anomalies you're going to see an uptick in revenue (assuming your marketing isn't totally brain dead).

I wish I had more fucking time to post on here, it really is the only OM forum worth reading (though for some reason I still browse Delhi Point). Unfortunately I've just been too damn busy working to hang out.

 

[eide]

For most cases, I think this would be a waste of time. Most websites don't get recurring traffic, unless:

1. They offer content / stuff worth checking from time to time
2. Website has good email marketing in place

In the first case, let's say blogs are typical example of such sites. If visitors are already willing to return, they are there for the content that interests them, so I think they just go to category or post that catches their attention

In the second case, custom landing pages for email marketing can solve the problem..

This kind of personalization is suitable for big e-commerce sites only I think..

 

[on-on]

For most cases, I think this would be a waste of time.

You're basically right. Most online marketers think in small terms, but almost all of them - even the successful ones - think in surprisingly narrow terms. If you're making a ton of money then why not? However, given those parameters, it wouldn't make sense to pay attention to any kind of persistence or data retention.

This kind of personalization is suitable for big e-commerce sites only I think..

No, it's suitable for anyone who is running a larger enterprise with a lot of persistent customer data. A number of online marketers who push a lot of traffic through their sites could do this, most don't, but that doesn't mean they're right - it just means that they're making enough money that they don't feel (or know) it's necessary (or possible) to diversify income streams and cross-connect data through a number of verticals and channels. I know WickedFire is all about people acting like Hip Hop stars, but I don't mean this insultingly, it's just human nature. If you're a "super affiliate" and you're making a very nice living, why work harder? We could all work harder, but we all hit a level where we're happy and there's nothing wrong with that. Once you have a large development and networking team to work with, though, you can do a lot of interesting things.