Tuesday, May 10, 2011

New way of monetizing the social information, a new ads paradigm (Part I)

Today the ads world is dominated by at least four paradigms. The second part of this posting will propose a new one called "CrowdAds" which naturally fits social networks and social sharing. Before that, let's review the current paradigms.

Google Adwords and Microsoft AdCenter are the undisputed champions of search ads paradigm. The idea is pretty simply. You submit your queries to your favorite engine and, together with search results, you will get some related ads. A whole ecosystem has been created where publishers can buy ads by using an auction mechanism. This process is now quite common, but I do remember old days when search was considered a commodity with many people claiming that there was no money in this business. Credit goes to Google for suggesting how to transform search into a cash-cow machine.

Google Adsense is the second major paradigm for ads. The idea is to embed ads related to the content of a given web page. This embedding leverages some algorithms for finding a suitable set of ads in a transparent and automatic way. This paradigm is largely used on Internet and allowed many people to earn money on the content they produce.

In-text ads is the third paradigm which embed ads related to the content of each web page directly in some artificially created links contained in the page itself. A popup window is shown when the user move the mouse over the artificial ads and this particular aspect generated many criticisms about the invasiveness of this paradigm.

Cashback rewards are websites which pay users for their activities. When a customer makes a purchase online, instead of visiting the retailer directly, they may choose to go via a cashback website in order to generate a monetary reward for buying the products or services.

My question to you is what is the best ads paradigm for social networks in terms of relevance, monetization and reduced invasiveness among the four one discussed above? Or would you suggest something else?

2 comments:

  1. Not sure where you would place coupons and deals, especially local and personalized coupons and deals, but I think those are badly underdone and a big opportunity.

    There's also a question of where you would put display advertising in your categorization. Not all advertising is judged on clicks. Some, even most, is targeting branding, not direction action, and is based on impressions and recall.

    Speaking of all this, did you see Paul Tyma's (creator of mailinator) new project, Clickochet? Worth a peek.

    http://mailinator.blogspot.com/2011/05/introducing-clickochet-free-ad-trading.html

    ReplyDelete
  2. I would suggest something else.

    All of the methods above (including deals) are delivery mechanisms which are algorithmic; that is, whoever runs the marketplace (Google, Microsoft, FB, Groupon) decides through an algorithm (behavioural targeting or click-prediction) whether or not the user is going to see a business offer trying to infer the current intent of the user (all based on past clicks and current keywords/clicks/click sequences). The problem is not whether or not the algorithm is correct but whether or not the user trusts the recommendation (as an example: how often have you bought a book as a result of an algorithmic recommendation Email by Amazon compared to a personalised book recommendation by a friend through Email - same text just a different sender?)

    An alternative is to become a marketplace of personalised recommendations between people in a social network (who already have a trust relationship!) The main difference to something like Beacon (which has been tried before and failed) should be the a-synchronicity, though: As a user, I am only interested to see a friends recommendation for a product when I am going to look for a product in the same category (with the intent of buying). Social networks provide all the information we need (who trust whom) and the technical means to bridge over time.

    Just an idea :)

    ReplyDelete