Commentary

Paying For Quality

If the old saying (usually credited to Ralph Waldo Emerson) about inventors and mousetraps were updated for online advertising in 2010, it would read something like this: "Build the right mousetrap, and only the highest-quality customers will beat a path to your door." In the early days of the Internet, advertising was about eyeballs, and advertisers were imitating the mass-market tactics used by their broadcast and print forebears. But breakthroughs in analytics have shed light on the quality of online traffic, changing the focus from lead quantity to lead quality. After all, what good are impressions if they're seen by the wrong people?

 

Google built its AdWords model to address this very notion; keywords are priced based on their demand in the marketplace.  In addition to testing different price points, advertisers can test different words, effectively targeting different users to see which ones produce the best conversions.  This is a useful model because advertisers are paying closer to what a customer is actually worth, but it's still a crapshoot: virtually every customer has a different lifetime value, and hence advertisers find themselves paying based on averages. An online seller of tennis balls might find that some of its customers are worth $3 and some are worth $300, even though each customer was searching for the same keyword  -- "tennis balls"  -- on Google. 

Having advertisers pay based on the number of eyeballs, clicks or acquisitions (preceding terms ordered by increasing measurability) is still suboptimal, as there is no real measure of "quality" built in. With standard Cost Per Acquisition (CPA), Cost Per Mille [Impressions] (CPM), and Cost Per Click (CPC) models, advertisers pay the same price for every broadly identified lead/impression/click, regardless of the lifetime value of the customer.

What if advertisers only paid for ads that brought paying customers?  And, what if advertisers were able to pay more for the good customers and less for the bad ones?

It's time for an online advertising model that, like Google's AdWords, takes advantage of new optimization technologies to provide easily measurable results for the right metrics: how do I get the right customer, and only the right customer, and how much does it cost?

PPQ: Pay-Per Quality

It's time to "pay for quality." In the offline world, advertising agencies are exploring pay-for-performance models (see this recent article from The Economist for more details) as a means to prove their value, but some are wary: "Some agency executives are skeptical about being paid for value, because it is so subjective."

But in the online advertising world, where transactions can be easily measured and traced through a series of mouse clicks, paying for performance or quality absolutely makes sense. There are online platforms now that can identify which traffic sources supply the most valuable customers, optimize the traffic for each advertising offer and adjust the advertiser's costs based on the quality of the lead. Using this new Pay Per Quality (PPQ) model, advertisers pay the right price for every lead by analyzing a customer's total lifetime value.  

PPQ In Action: Alternative Payments and Online Advertising

The first implementations of PPQ advertising have been in conjunction with ad-supported payment systems. Increasingly, online retailers are embracing alternative payment systems like PayPal and Bill Me Later to help reduce shopping card abandonment by streamlining the payment process. Other alternative payment platforms merge advertising and payments with "ad-supported payment systems." In this scenario, customers "pay" for Retailer A's item (in fact, they get it for free) at checkout when they complete an offer presented from Advertiser B. These offers from advertisers are highly contextual and can be targeted for each customer based on a variety of criteria.

On the surface, the benefits to customers (a free product, service or gift card with their purchase), retailers (increase their customers' willingness to pay and make more sales without expensive processing fees) and advertisers (highly targeted leads with a high lifetime value) are obvious.

But digging a little deeper, you can see that advertisers are actually paying for quality rather than quantity -- as certain alternative payment platforms use optimization technology to present the right ads at the right time and price leads on a pay-for-performance basis. Some alternative payment platforms include intelligence that recognizes that each customer is worth a different value.  Advertisers then work directly with these payment providers to identify which traffic sources supply the most valuable customers, and the payment provider adjusts the advertiser's acquisition costs based on this "quality" data.

These trendsetting advertisers are closer than ever to their industry's Holy Grail: paying only for ads that lead to high-quality customers.

4 comments about "Paying For Quality".
Check to receive email when comments are posted.
  1. Alisa Bartholomew, January 18, 2010 at 5:01 p.m.

    Are there commission-style pay-per-quality ad models? For example, for advertisers that quantify the quality of the customer as $ revenue, could they pay a % of that final purchase amount? (i.e. The customer transaction resulting in a $300 sale is worth more than the $3 sale so the ad price is a percentage of the sale.) A customer who results in multiple $3 sales would still yield a higher price ad because the overall revenue/lead quality would be higher).

  2. Carolyn Allen from California Green Solutions, January 18, 2010 at 5:16 p.m.

    As a publisher who "educates" people before they are ready to buy... I have a problem with all the advertising credit / money going to the LAST impression. It takes numerous cumulative impressions for a person to get ready to buy. How do WE get paid for that process? I'm beginning to think that publishers need a "tip jar" to compensate them for all this early demand generation cultivation.

  3. Eric Melchor from Smart Digital Spending, January 20, 2010 at 4:42 a.m.

    Even without using Google technology, a similar approach would be to use the 'weighted point' method - establishing a numerical level of importance for key pages on your site. This approach can be used for both paid search and online display banners. OnlineMediaAnalyst.com provides a free tutorial on how to setup this measurment approach at http://bit.ly/17fqh9
    Eric
    SmartDigitalSpending.com

  4. Richard Frankel from Rocket Fuel, January 21, 2010 at 9:14 p.m.

    Excellent article Alex. The good news is that there is an emerging solution space that provides exactly the kind of PPQ objective that you are looking for.

    The latest ad exchanges (Google's AdExchange, Appnexus, Pubmatic, etc.) are beginning to provide for what is called "real time bidding" or "RTB" solutions. This allows buyers of ad views to bid on each impression opportunity that shows up, one at a time, in real time.

    The hard part in this is knowing WHAT to bid (not to mention all the attendant technology that allows for such bidding). Companies like mine are building model-based platforms that effectively predict the likely value of each impression and can bid accordingly.

    As these systems scale up it will be a boon for advertisers and publishers alike. Advertisers will be able to buy much more efficiently, and publishers can get paid full value for their audience. It should be better for consumers as they get shown ads that are more relevant to their interests.

    Other problems (such as fair attribution) remain to be addressed!

Next story loading loading..