There are three different feeds you can subscribe to, which correspond to the different types of content. Links are short posts of one or two paragraphs which simply share an interesting link with brief commentary. Entries are longer prose pieces of at least 300 words. If you subscribe to everything, you'll get everything I publish, including approximately one entry and five links a week. Comments are not included in the everything feed. If you want to subscribe to comments, you can keep up with the conversation using the comments feed.

Timestamp
Published at 12:02 AM on September 25, 2009
Comments
There are 5 comments on this post. Add yours.
Content

My friend Joe Thibault recently emailed me about a new site he has started with Sean Behan (a crackerjack Ruby programmer). Their site, called PostLearn, is essentially an affiliate job board for education. While I wish them success in their endeavor, I won’t be joining as I believe this model suffers from a couple of major flaws.

PostLearn affiliate graph

Right: Graph of PostLearn affiliate traffic, lead by freetech4teachers.

The greatest flaw is immediately visible from the affiliate graph: one site drives a vastly disproportionate amount of the traffic; Free Technology for Teachers alone accounts for over 90% of the traffic, and consequently will receive far more affiliate revenue. Except it will probably receive all of the revenue. In an affiliate model, there is a certain base amount of traffic required before you can get a single sale, a minimum that I doubt many of the others on the long tail (including yours truly) would reach. Effectively, the entire pot of affiliate money is controlled by a tiny oligarchy of sites.

This would be acceptable (it’s a free market, after all) were it not for the fundamental flaws in the Internet economy. This flawed economy is controlled by a small oligarchy of noisemakers. This oligarchy isn’t particularly hard to enter: just abandon journalistic ethics and post lists of the top n ways to write lists.1 Boom! Traffic.

Unsurprisingly, just as in the larger Internet, the most popular PostLearn affiliate simply rehashes merit-less news stories and tips.2 Looking down the page, lo and behold, we find a list of the aforementioned variety. Sadly, this content is rewarded far more generously than potentially more deserving comment.

Google AdSense grants us some slight freedom from this paradigm by democratizing advertising. Though the most trafficked sites still receive the vast majority of the revenue, smaller sites do share in some of the revenue, potentially enough to offset minimal publishing expenses. In this way, Google AdSense resembles most modern democracies: while the elite still maintain most of the power, the little guy does get a small voice. It’s not perfect, but it’s better than an oligarchy of the unthinking elite.

Fusion Ads

Above: The exemplary Fusion Ads logo.

However, we can do one better with a meritocracy. Rather than being a detriment, advertising can be a force for good on the web. Fusion Ads and The Deck are prime examples of this. Instead of rewarding breadth of content, these networks focus on finding blogs which consistently post material with depth. By handpicking their members these networks ironically end up leveling the playing field and giving quality content a chance to shine. I truly believe this is the advertising model that will save the web.3

With that in mind, I offer this challenge to Joe and Sean: devise an advertising platform which will improve the quality of the edublogosphere. In my opinion, such a model should observe three crucial principles:

  1. Popularity doesn’t necessarily correlate with merit. Meaningful blogs with strong reader relationships will, in the end, provide greater long-term benefit.
  2. Be selective. The quality of small, niche groups is more easily controllable, making them more marketable.
  3. Know the audience. Most readers are going to be job seekers (teachers) not job posters. The billing and payment strategy should reflect this.

Whether they choose to implement these ideas or not, I wish Joe and Sean the best of luck and look forward to their response.

Notes
  1. Paul Graham, one of the few who has an audience and integrity, wrote about this phenomena in a must-read essay.
  2. This isn't a condemnation of the site's manager. If he is happy with poor design and meaningless content, that's his decision.
  3. No, I'm not a shill for Fusion, just a fan.

Comments

There are 5 comments on this post. You can add your own below.
Content
Your writing style and content both leave me astounded. Excellent post. Superb.
Name
Content

@Abraham: Thanks for your kind words, both here and on Twitter.

Content
Thanks for the review. I wouldn’t take too much away from that graph. Freetech4teachers.com has the majority of the traffic, but we’re so new that it will take a while (and more affiliates) before there is any meaningful data available. It was more out of curiosity that I visualized the data and less to give prospective customers a sense of our reach, which is very small at the moment. Our affiliate partners are most likely aware of the traffic they are pulling in and how this might translate to revenue. I am surprised that you think Google Adsense has a better (or more meritorious) model. I see Adsense as responsible for many of the aggregate, spamesque web sites that attempt to capitalize on meaningless clicks. In our model clicks and impressions are irrelevant, instead we focus on conversions. Affiliates that can deliver conversions get rewarded. It is our thinking that this will enhance the quality of the service, without making any judgments as to the quality of affiliate content. Our only requirement is that affiliates are within the realm of education. Again, thanks for your review. I appreciate your input and both Joe and I will take some of the points you raised as we improve the product.
Name
Content

@Sean: While the graph probably isn’t representative of the overall affiliate market (or what PostLearn will hopefully attain), I do think it is indicative of the general problem. While the curve isn’t necessarily always that dramatic, the long tail is still a fundamental part of the web. Unfortunately, those at the far left of it aren’t necessarily the most deserving.

On smaller sites, especially those using networks such as Fusion, the click-through rate can actually be higher than on larger sites: readers are more engaged with the content and what surrounds it. I firmly believe that niche, high quality advertising of the web.

The main problem I have with affiliate models is this: they require a massive volume of traffic to attain any reward. Lets generously assume that 1 in every 500 people going to PostLearn will end up purchasing a posting (feel free to prove otherwise with data). Then lets assume that 1 in every 100 visitors to a blog clicks through to the site (again, generous). So, only 1 in 50,000 visitors ends up making the affiliate money. Before I can make even $14, I’d have to get 50,000 visitors. Sure, my site is admittedly small, but I don’t even get that in a year.

Contrast that with Google AdSense. Given our previous 1/100 click-through-rate and a cost-per-click of $0.50 we’d receive $250 for those 50,000 visitors. Even for my lowly 1000, I’d get $5—far more than nothing. Google AdSense scales very well: everyone makes something. Affiliate programs tend not to: small sites essentially provide free advertising with no payout.

In terms of concrete improvements you can make, I think your payment system should be tweaked a little. Maybe look at how Authentic Jobs works. Essentially, payouts should be made to affiliates when someone responds to a posting from their site. The widget would rotate through recent postings. Remember, the principal audience on affiliate sites is generally teachers: not administrators. Having this rotation can significantly improve your sales pitch: you’re now selling space on all affiliate sites.

Content
I appreciate the review…I just want to make a few clarifications.

1st, I think you discount the value that Richard Byrne @ freetech4teachers.com provides to his audience. Richard churns out a LOT of website and resource/tool reviews (he helped us get our first 5 affiliates by posting about us). His site represents an aggregation of the newest and best free resources to teachers (it’s why he is generating a lot of traffic: he provides resources to all subjects and levels of educators but also gives a quick synopsis of how he would use it in the classroom). Quality, especially on the internet, isn’t exclusive to original content.

2nd, I think you misinterpreted Paul Graham’s post about lists. He never suggests that they are inferior. In fact, he rather suggests that they may be a more efficient and succinct way to provide information to readers (which is why audiences love them and writers love to write them — they’re easy to create and digest and have value even if you disagree with one/some/all of the points. Note that THIS is an N list of my points…it doesn’t mean I didn’t think about it critically). Though I’ll agree that the "100 best __" lists have little merit. Paul also goes on to suggest that typical expository writing (where there is an intro, N points as the body and a conclusion) is a waste of words. Why not just focus on N points?

3rd and final: in your comment to Sean you suggest we tweak the payment system and look at Authentic Jobs, that only reaffirmed my belief that Postlearn.com can and will work. They are exactly the same model.
PL: $65 | AJ: $250
PL: provide payment only for a post conversion | AJ: provide payment only for a post conversion
PL: targeted affiliates (edu) | AJ: targeted affiliates (web/creative)
PL: 20% payout | AJ: 30-33% payout

Note that we already rotate live jobs through our widget; providing a fresh. This discourse did get us thinking though…so look for some nifty enhancements shortly from http://postlearn.com.

Thanks again for sharing your review!

Respond

Respond to this entry with your remarkable insights.

(Required)