You’re an affiliate marketer or an affiliate network operator. You deal with massive amounts of data and data relationships every single day. You often find yourself not only having to receive certain data but having to send back data to your partners, post conversion.

The passing back of data to your partners or clients might seem pretty rudimentary, but you’d be very surprised to find out that many tracking systems lack a comprehensive version of this functionality because of variable scope issues.

Variable scope is basically a set of rules that defines where one can and can not access a variable. About.com explains it best here, “you can’t define a variable at any spot in the program and use it at any spot, there is a specific relationship between where a variable is defined and where it can be used.” What this means for your tracking platform is that any data that is required to be ’sent back’ to your partner/client through a pixel or any other means must be within the ’scope’ of that end-pass or end-population.

Example:
- A publisher sends a value to you through your tracking URL and needs it passed back to their pixel.
- If this value, passed into a parameter in the initial URL, is definable and in the scope of the pixel, you can pass it back without a problem.

I won’t go into specific implementations here, because the definition and handling of local and global variable scopes varies from framework to framework. The logic, per the above, however, will never change.

If your current system doesn’t allow much flexibility for data handling, you may want to consider switching software!

I recently came across a situation in which a line of dynamic, custom text had to be written to a screen / file for every IP address within an IP range. At first, there only needed to be lines for 10-20 records at a time, but that soon changed. Before I knew it, I was challenged with the same scenario, only with thousands of IPs this time.

Here’s some pseudo code to help you code up a pretty turn-key solution, using simple nested loops and iterators:

// For the ENTIRE IP Range, AAA-BBB, output the following, using the syntax:
// XXX.XXX.XXX.CurrentRange (plus any other info you’d like, including any sub-iterated data)
// XXX.XXX.XXX.CurrentRange (plus any other info you’d like, including any sub-iterated data)
// XXX.XXX.XXX.CurrentRange (plus any other info you’d like, including any sub-iterated data)
// XXX.XXX.XXX.CurrentRange (plus any other info you’d like, including any sub-iterated data)

// Main Loop
IF AAA != BBB {
CurrentRange = AAA;
Output to text area : XXX.XXX.XXX . CurrentRange . any other info you’d like . “\n” ; // where ‘/n’ indicated a return to the new line
AAA++; // increment your position in the IP range

// Iterator Loop
// An optional loop should you wish to insert another set of sub-dynamic data or information that does not occur in the linear, incremental manner of the parent loopTOPVALUE = yourCielingValue;
IF subIterator == TOPVALUE // where TOPVALUE is the value at which you should start over at zero.
// subIterator can be any numeric data type. An array is probably best suited for this situation.
subIterator = 0;
ELSE
subIterator ++
}

This structure is very basic, but very useful. You can use this same pseudo code to crawl through any set of values with a ceiling value and perform a desired writing to a file or web field on a per-record basis. Let me know if you have any questions or if there’s something else you’d like to see here.

Linux, per many of my other blogs, is truly the King of operating systems to couple with any stable email marketing endeavors.

As an email marketer, you will often find yourself having to locate opt-in information or pinpoint email records quickly. Here’s a quick tip on how to find a record really fast using my favorite text utility, VIM (don’t copy the quotes):

- Open your file by running the command:
“vim fileNameHere”
- Search for your text by executing:
“/email@address.com/e”
- The above text searches for the string, “email@address.com”. The ‘/e’ at the end makes your cursor land on the found record to make your data hunt even easier.

Special Note :: VIM is, of course, available on Mac OSX, as it runs on a Unix core, bundled with quite a bit of shell/terminal applications!

That’s it! You can search for any string very easily using the above snippet. Questions? Email me.

With so many tracking systems dependent on cookie tracking, how can you still track consistently without having to rely on the end-consumer’s local browser settings? Don’t always use cookies.

To track without cookies (and accommodate almost any browser configuration/environment), one must take a look at what cookies do. This is a bit over simplified, but it works like this (I’ll be using an “opening/closing door” analogy):

- Consumer clicks through a URL
- Cookie is installed on local machine, opening the door
- Consumer performs any actions (reading the page, making a decision, etc)
- Consumer performs the CPA ( Cost-Per-Action ) submission
- “Thank You” page, also known as the “Confirmation Page” is reached, which houses the tracking pixel
- The cookie session is closed, telling the tracking system a “sale” or “action” has been made and the door is closed.

If you break this down, the linear nature of such tracking shouldn’t have to rely on cookies. In Linktrust 6.0, there is another option to cookie tracking called “ClickID” Tracking. This form of tracking involves the appending of data onto a URL and then later onto a pixel. So, the set up is the same as cookie tracking with these added values being the exception. This form of tracking still tells the tracking system the relationship between the inbound click and the conversion, or sale or payable action, that was made. Because the value of each ClickID is unique, this form of tracking is considered to be secure.

Questions? Email me!

Does your current tracking or ad-serving platform not offer you click analytics as a feature? While that is a large issue that should probably be addressed at the platform level, you still have options.

Since the operator of the given non-click analytic-ready platform controls all of the inbound links to the system, you can create a ’step’, or ‘agent’, in between the user’s click and the user’s landing on the offer page. This will then serve as a sort of ‘extension’ to your tracking platform, leaving the customization functionalities completely up to you. All that really needs to be done is the integration of a simply click analyzer, which can be any PHP/PERL/ASP script that records the click’s origin and time (and if you want to get really advanced, which conversion, if any, a click is associated with).

Set the ‘click agent’ to record the necessary info to a file or database, then set it to redirect to the appropriate page. This means that your script will need to receive a few values, not necessarily recording them, that identify the : creative, affiliate, campaign, sub ID, etc. Your script, after recording its information, will then continue to route the traffic to your client’s tracking URL, passing the parameter values to the next URL. The ‘click agent’ URL that could be set up in your system could look like this:

www.myClickAnalyzerDomain.com/scriptFile.php?campaignValue=xxxx&affiliateValue=xxxx&creativeIdentifier=xxxx&SUB=xxx

Since runtimes and page redirection times are one of the key players in conversion rates, it is important to optimize any script or extension you write for your tracking. Here’s how I suggest you optimize your click analytic tool:
- Use distributed systems
- 1 system or process will record the click data and perform any necessary routines
- 1 system or process will send the redirection forward

The idea here is that redirection will keep going while the other processes complete independently, not hindering runtime in any way.

Have any questions? More ideas to make your click analytic utility even better? Drop me a line

If you’re an email marketer, you love data. It is the blood that flows through your ESP’s (Email-Serving-Platform) veins and the life essence of your success (in addition to your deployment, templating and configuration skills, of course).

Chances are, you probably get a few daily data feeds, but in this blog, we’re talking about Legacy Data, or past data that has been collected into a larger-sized list or file, generally speaking.

The benefit of Legacy data can be seen in the sheer sizes of the data packets that are received and sent-to. More data means more volume capacity, which then translates into more revenue.

Because the data in Legacy sets is compilation of records, whose opt-in dates originate from months’, or even years’ past, there are a few problems that can arise. First, complaints may arise. The older the data, the higher the chance that the end-consumer has forgotten that they signed up for a given mailing stream or newsletter. This can yield a larger volume of complaints, generally stating that the consumer doesn’t remember signing up for the mailings, deeming them unsolicited and “Spam.”

Another problem that can occur when working with Legacy data is that there can be a large percentage of messages per data set that doesn’t “exist” any more. The more amount of time that has elapsed between your drop date and the date of your data set’s opt-in, the larger the likelihood that the consumers on record have changed or closed their email accounts.

The first issue is unavoidable. All one can really do is address the issues in a 100% Can-Spam compliant manner and provide the associated opt-in (or sign up) information to the parties involved. The second issue can be remedied by implementing utilities that scan your data for validity, pre-drop. I have personally not seen a utility that can do this, efficiently and preemptively scanning millions of records.

I think that would be a great idea if it hasn’t been engineered already. Let me know if you’ve heard of such software!

Have a great week!
Josh

I know I’m going to get a lot of heat from those of you who are addicted to your iPhones, but I had to shine light on one of the best, most functionality-packed smart-phones that I’ve come across. I just got my BlackBerry Bold and I must say, I absolutely love it. Let me count the ways…

- Stats (and webpages) are more accessible than ever. I logged right into Linktrust and the ASP pages looked great after my mobile browser resized them. I also logged into a few other business associates’ tracking systems and they were scaled perfectly for my screen. If you want stats on the go, without having to use a supplemental 3rd-party application, this is great.

- Scripts. You heard it right. Recently, I wrote a script in my text editor on my Blackberry Bold and uploaded it to the server on which it needed to go all from the little powerhouse in the palm of my hand. A year ago, that would have been crazy-talk.

- You’re probably wondering how I uploaded this to the server. Well, the BlackBerry bold supports all sorts of protocols right in the default browser (including FTP). I was actually able to download a 3rd party SSH application and get into the server to perform all of the operations and file transfers that I needed. *Side note : never SSH into any server as root from any handheld device. That is a horribly bad idea.*

The Blackberry Bold also comes jam-packed with a plethora of other features such as WiFi over 3G integration, real-time video streaming (and the buffering is nothing to shake a stick at), a great camera, and an awesome camera, to name a few.

If you’re in the Affiliate Marketing industry, or any technology-oriented industry for that matter, the Blackberry Bold will be an asset to your success on-the-go.

I recently had to help one of our prospective merchants set up their systems to ensure that they displayed pixels  only once per unique transaction.  The situation was one in which it was possible for the end-consumer to refresh or revisit the page that the pixel lives on (their confirmation page was not a “thank you” page, but rather a second step in a lead-gen campaign).

Per my conversation with this associate, here’s how you implement the solution by ’surrounding’ the pixel with a conditional wrapper.  The logic (written in pseudo-code below) should work like this:

IF Unique Transaction ID
Display pixel with OrderID
IF NOT Unique Transaction ID
Do NOT display pixel

This solution ensures that regardless of how many times an end-consumer visits a page (for any reason), the pixel will fire only one time, as it should.  This can easily be implemented in PHP or Javascript using simple, nested ‘IF-THEN’ loops.

Questions?  Feedback?  Email me!

If you’re an emailer, you are already familiar with suppression and opt-out management services provided by UnsubCentral (click here to visit their website).  You may have noticed that instead of the usual process, which involves downloading readable text-based suppression files to scrub your send-to lists against, they have implemented a slightly different process involving encryption.

This method of suppression works like this:
- You visit the advertiser’s Unsubcentral suppression management page
- Download the MD5 encryption utility and encrypt your list.
- Compare the encrypted lists on your machine
- The resulting list, with a pair of unencrypted records, is the cleaned list.  (you’ll just strip out the delimitated column that’s not encrypted).

(See a more comprehensive breakdown of the process HERE)

The only issue with this is that, although it takes a lot of resources and runtime to complete, MD5 is broken and can be decrypted (albeit record for record, which would take a reallllly long time to decrypt a sizeable amount of data).

I spoke with Josh Janicek at UnsubCentral yesterday, and he informed me that they are making a move very soon to offer  SHA-256 encryption as an alternative to MD5 encryption, to ensure clients’ lists are protected.  This alternate encryption algorithm has been declared safe by the NSA for the time being.

What are your thoughts on encryption and suppression processes?

So you’re an Email Affiliate Marketer, aka, “Emailer.”  This coveted role comes with a lot of, well, data.  Data, the key to any email campaign or mailing, is abundant in the email marketing world, and the manipulation of said data is required, whether it be necessary to format, merge/split, or remove any unnecessary characters from therein.

Windows and Mac systems can be beasts, but depending on your configuration and the size of the data you wish to manipulate, they can prove useless against even the simplest ‘find and replace’ task.  My advice as to the best solution?  Use Linux.

Even the lowest-level hardware configuration in Linux does a better job at manipulating data than the most decked-out Windows box using standard file utilities.  Linux comes out of the box with every data-manipulation capability under the sun that an email marketer would need.  Its wonderfully compact and efficient algorithms make even the most daunting task doable in a reasonable amount of time.  Just use the VIM, VI, or even EMACS utilities.

Want to know more about the benefits of using Linux in the affiliate marketing space?  Drop me a line!