What Job Does Your Affiliate Site Do?

When you construct an affiliate site or an affiliate program or look to optimize a site or program, the question of how and why people visiting your site or program is just as important as what they are doing on your site. Both affiliates and advertisers frequently overlook the essential question of what job people are looking to accomplish by searching for your keyword or discovering your site.

On the way to a meeting in Charlotte yesterday, I got to listen to a couple of podcasts I’ve missed. One of those was an episode of The Critical Path and it might have been the most revelatory thing I’ve heard in a long while on this very topic.

In a nutshell, Horace Dediu talks with a guest on the fascinating concept of Jobs To Be Done.

The concept is deceptively simple… people looking to buy a product or use a service (go read the Facebook post linked above) don’t actually buy the product or service as a thing, rather the customer is looking to hire the product or service to accomplish a job for them.

The example of a Snickers bar vs a Milky Way bar in the podcast makes it clear. You should seriously stop what you’re doing and go stream/download this now…

5by5 | The Critical Path #19: The hiring and firing of milkshakes and candy bars: “Horace talks with Bob Moesta, a pioneer of Job To be Done research. We go over the theory and process of understanding what products are really hired to do and ask why this understanding is so hard to come by. In a discussion rich with examples from multiple industries Bob illustrates how marketing, design and engineering are all dancing around the question of how product should be developed. Could the universally accepted compartmentalization of corporate functions be a root cause to product failure?”

Here’s the mp3 or click over to hear the stream (and subscribe to the podcast and the other great shows on the 5by5.tv network that you should be listening to!).

Namecheap Domain Move Update

Yesterday I decided to go with the gestalt and move a few test domains from GoDaddy to Namecheap.

I’m happy to say that at about 9pm last night, I received an email from Namecheap saying the process had been completed (and one from GoDaddy saying goodbye).

In total, it took about 10 hours from start-to-finish, but I’m guessing much of that was due to the Dump GoDaddy Day movement.

All of the name servers were switched over automagically (they point to my host MediaTemple for the samharrelson.com domain) without me having to do anything, which is a big relief.

I’ll be moving PayPerTrends, as well as all of my domains, over to Namecheap this weekend. So, if you see anything goofy or the site has some downtime, you’ll know the hamsters in the back are doing their thing.

Excited.

I’m a nerd.

Rock on.

Facebook Like-Jacking and Need for Digital Literacy

Interesting:

Criminals Used Affiliate Marketing Sites in Majority of Facebook Scams in 2011: The vast majority, or nearly 74 percent, of Facebook attacks in 2011 were designed to lead users to fraudulent marketing affiliate and survey sites, the report found.

Affiliate marketing was a “rich source” of income for scammers, according to Amir Lev, CTO of Commtouch.

First, it’s interesting to me that the writer focuses so much on how easy it is for scammers and “criminals” (a conviction is needed to be a criminal… just saying) to use the medium of what he broadly labels as affiliate marketing. The piece focuses more on survey type deals that were so popular with the “free iPod” craze of 2003-4 in the pre-CANSPAM era.

It’s pretty easy for the legitimate businesses he sources as being defrauded to check their logs and any affiliate manager or OPM worth their salt will catch this kind of scam traffic, especially if they are dealing with the lead based side of things in the CPA and lead gen areas.

The real heart of the piece should be about the need for better digital literacy among users of spaces like Facebook (especially if they are browsing on a Windows machine with IE6 or 7).

Cue Wayne Porter

“For criminals, it was not enough to just trick users, as criminals need to make sure the attacks spread and continue to trap other people, Commtouch said. They were most likely to trick users into sharing the links almost half the time, but also tricked users into copy-pasting malicious code to trigger a cross-site scripting attack or downloading malware. Rogue applications and “like-jacking”—which employs a malicious script on the page to convert any mouse clicks on the page as a “like” that is also visible to other users—were employed in about a third of the scams.

“In 48 percent of the cases, unwitting users themselves are responsible for distributing the undesirable content by clicking on ‘like’ or ‘share’ buttons,” according to Commtouch.”

It’s fascinating to me that many of the conversations Wayne and I were having back in 2008 about a future of social-engineered badware that would find virility through good-willed sharing are coming true in 2011 and even more so into 2012.

At the root of the issue isn’t affiliate marketing or how easy it is to scam businesses. Businesses have failsafes and checks in place to catch these things (ideally). Instead, we need to have more savvy users who realize the implications of sharing or liking a suspect link or article or site.

This sort of manipulation of otherwise trusting, naive or uninformed users of the web will only intensify as more people go on the web with mobiles and tablets in the coming five years.

The Problem with A/B Testing Ads in Social Media

Something to remember as you repeatedly hear “test test test!” and embark on your own split testing for various creatives via Google Analytics etc…

How Not To Run An A/B Test: “Although they seem powerful and convenient, dashboard views of ongoing A/B experiments invite misuse. Any time they are used in conjunction with a manual or automatic “stopping rule,” the resulting significance tests are simply invalid. Until sequential or Bayesian experiment designs are implemented in software, anyone running web experiments should only run experiments where the sample size has been fixed in advance, and stick to that sample size with near-religious discipline.”

Basically, don’t peek at your testing, don’t test for significance and have sample sizes in mind for your tests.

This is insanely important in more emerging areas for creatives like social media or mobile.

This looks like calculus but it’s a good reminder that the observer often influences the test.

Yay science.

via Hacker News