News out from Google today that they are expanding their Pay Per Action beta on a global basis.  (Full release here).  Although now world-wide, the PPA beta is still limited to the AdSense publisher network and doesn’t touch Google’s own paid search listings.   

As I mentioned in my March post discussing the Google beta, Google has a huge cash cow to protect in its Pay Per Click advertising system, and its move towards Cost Per Action advertising would likely be slow and methodical.  This news didn’t disappoint. 

As long as the PPA program at Google is limited to third party publishers on Adsense, this program will have limited success.  As a publisher, it is far easier for my site to generate click revenue than conversion revenue.  What’s more, Google’s PPC rates are extremely high because of the liquidity of its PPC back end auction.  Google will need that kind of competition in its PPA program to create higher rates as well over time.  

Google is asking its publishers to take a lot of additional risk, something that I’m sure the vast majority of them will be unwilling to do until it is clear that higher ad revenues await.         

Internet Pay Per Click search advertising is getting complicated.  Our CEO often calls it Internet Chemistry, but it’s really more like a form of Actuarial Science that is increasingly applying complex statistical methods to predict future advertising performance (e.g. the likelihood of a conversion).

Niki at Bronte Media was recently discussing an interview with Anil Kamath, one of the founders of Efficient Frontier, a search marketing firm.  Anil’s interview is a great example of how PPC marketers are becoming more and more like hedge fund managers on Wall Street.  Why do you need a PhD in predictive analytics to be a PPC marketer these days?  Two main reasons: 1) the tremendous growth in the long tail of available keywords; and 2) the giant chasm that still exists between the click and the conversion. 

Basically, Efficient Frontier and other PPC gurus try to optimize search ad campaigns by analyzing click and conversion results from a really big data set of customers.  Here is an excerpt from Anil’s interview:

You can’t just Bid a keyword based on its own conversion; you need to look at the elasticity of the Bid as well in terms of how your bid effects your position in the market place that you are participating.  What the competition is bidding against you, and how many clicks you’ll get more or less based on whether you bid more or less on that keyword.  When you take the information of that keyword, and combine it with information that you have about other keywords; you’ll have a pretty complex problem. 

Complex indeed.  It seems to me that statistical expertise like this is the reason that Nextag-the master of using large sets of keyword data to create search arbitrage opportunities-was recently valued at $1.2 Billion. 

But does search marketing really need to be so difficult?  Efficient Frontier uses a lot of complex statistics to back into some pretty simple overall business goals on the minds of direct marketers.  In Anil’s words:

The business goal could be to spend a hundred thousand dollars and maximize my registrations, or maximize my revenue.  Or, it could be to get twenty percent margin, while getting you the most revenue possible.

Contrast this to the simplicity of Cost Per Action search advertising in general, and the Jellyfish Value Per Action model in particular.  The beauty of VPA is that it closes that chasm between the click and conversion, eliminating risk and a lot of math homework in the process.  Marketers’ jobs become much easier when they only pay for successful conversions because they can easily translate their ad spending into their overall business goals without a lot of complex math.  At Jellyfish, we have lots of advertisers now using our back-end ad platform to manage their CPA commissions at a product and category level.  And each time they set their commission level, they are directly setting their rankings in our system and the return on their ad spend, without having to worry about click burn, the potential for negative ROI, or the complex statistical analysis of the PPC system.   

This root simplicity (complexity will develop in CPA-optimization land as well) is one of the reasons that we think Cost Per Action advertising will continue to grow to become a dominant form of online direct response advertising.  Here’s hoping Jellyfish will be a big part of it. 

Jellyfish is pleased to announce a partnership with Channel Intelligence today that will help accelerate our efforts to add new retailer partners to the Jellyfish.com shopping search engine.  Full Release here.    

This news is an opportune time for me to talk briefly about the Jellyfish advertising model.  There has been a flood of new shopping search engines over the past several months, but to my knowledge, Jellyfish is the only shopping search provider that is doing the hard work of building its own Cost Per Action advertising platform.  (Please correct me if there are others).  Granted, there are a handful of shopping engines that utilize affiliate providers such as Linkshare to integrate CPA advertising into their mix, but these big affiliate providers have illiquid commission structures (typically a standard store-wide commission rate for any sale).  

Why are we building our own CPA ad platform?  It’s a key part of our long range vision for a more consumer friendly version of CPA advertising that we call Value Per Action, or VPA.  Retailers in the Jellyfish system can adjust their advertising rates on a global store level, product category level or at the individual SKU-level for each product.  This makes perfect sense, as retailers may be willing to pay more or less in CPA ad fees for a sale depending upon the product being sold.  

Probably the best way to consider the advantages of this platform is to compare the Jellyfish model to Google’s remarkably successful Cost Per Click keyword ad platform.  Through competition, Google has created a liquid marketplace of Cost Per Click ad rates on millions of keywords.  Every day, online retailers compete with each other to achieve higher rankings on keyword searches by agreeing to pay Google more for each click, creating higher and higher advertising fees for Google.  Over the past several years, the true market rate for a click on keywords such as ”nikon digital camera,” ”mp3 player” and millions more have been set by this efficient system.   

With the Jellyfish CPA ad platform, online retailers bid not on keywords, but on the actual products they sell that match up in our shopping search engine.  The more stores are willing to pay for a successful sale, the higher they go on our rankings.  But here is where our unique model creates a more consumer friendly version of CPA:  In our VPA model, we always share back at least half of those advertising fees with the end consumer.  Thus, as retailers compete for search ranking by increasing their CPA commissions, BOTH Jellyfish and the end consumer benefit directly from that competition.  It’s a liquid CPA marketplace that creates lower prices for end consumers.  (you can read a full description of VPA here)  

For Jellyfish to see this vision through, we need two things: 1) lots of people buying products through our site; and 2) lots of quality retailers taking advantage of our risk-free CPA model to compete for those sales.  Channel Intelligence is a tremendous partner to help us achieve goal number 2.  CI customers include nearly 200 of the world’s best known retail brands, all of which now have direct access to the Jellyfish platform.  We look forward to helping CI and its clients take full advantage of the accountability and precision of Jellyfish shopping engine.                      

Brian Quinton at MultiChannel Merchant published a great article yesterday on the Google CPA beta test (article here).  In addition to some of the same observations I have made previously (here), Brian highlights a few additional potential limitations to Google’s CPA initiatives that I think are worth mentioning.  These include:

• Advertiser Concerns Over Sharing Transaction Data with Google.  Several ad industry experts quoted in the article highlight this as a concern, especially for large clients.  Because Google often controls such a large share of their online advertising wallet (non-CPA based), the advertisers don’t want Google to know their actual ROI on that advertising, out of fear that this will lead to higher rates, more competitors, etc.
• Double Counting Conversions.  Conversions may take place days or even weeks after a user clicks on a Google CPA ad and picks up a Google tracking code.  But what happens if that same user also accessed another tracking link from a third party affiliate network for the same offer prior to the actual conversion?  The result is that the advertiser may get charged CPA ad fees from both Google and another CPA network for the same action.  This is one of the things that I love about the Jellyfish model, since our customers have a built in incentive (through cash back savings), to ensure that they utilize the Jellyfish CPA link immediately prior to their purchase, creating less likelihood of a double counting issue.  

The article is worth the read for anyone following Google’s CPA test.      

Much has already been written about Google’s news last week that they are opening up “Pay Per Action” advertising opportunities within their AdSense network (Google Post here).  After digesting the news and the blogosphere discussion, I have the following observations:

This Won’t Change the World Overnight

• Google has a huge cash cow to protect in its Cost Per Click revenue model.  But CPC advertising suffers from some growing problems (e.g., click fraud and search engine spam that will ultimately diminish its power)   Thus, Google is forced to move towards CPA advertising (which helps eliminate these problems), but I’m sure that Google’s moves will be slow and methodical.  This news is a case in point: the new CPA system is being rolled out in beta and is limited to its AdSense network of third party publisher sites (where click fraud and spam are at their worst).  But how quickly will advertisers, and most importantly third party publishers, move towards the model?
• Bad actors in the Adsense system are not going to adopt CPA unless there is an equal opportunity to defraud the CPA model within AdSense.  If a publisher has built a network of click optimized sites, or developed click fraud practices that allow it to fly under the radar, I doubt that that publisher is going to jump at the Google CPA option.  A number of comments in the TechCrunch post highlight the fact that Google’s CPA system will allow a number of different kind of actions short of an actual sale that might create new kinds of “Action Fraud.”  So if the scammers do move over, it won’t help advertisers or consumers.  This is just one reason why Google is also pushing its Checkout product to control the shopping cart and measure conversions with precision. 
• Legitimate third party publishers are being asked by Google to assume a significant amount of additional risk.  With CPC, publishers control the revenue generating activity (the click) on their own site; with CPA, that activity moves over to the end advertiser.  For publishers to jump in, they are going to have to see significantly higher advertising rates and a highly integrated relationship between advertiser and publisher that helps the publisher qualify traffic for conversions.  Google’s CPC system has thrived on a low touch, automated system.  Again, several of the TechCrunch post comments highlight this disconnect.  Google will need a long term, sustained effort to convince publishers that this added risk is worth it.  I suspect this will take several years.      

Google’s CPA System Will Create Some Benefits for Advertisers and Consumers

• Jellyfish was the first comparison shopping engine to adopt a 100% Cost Per Action ad model, so we are happy to see a giant like Google help to push mainstream adoption of CPA.  As Google’s CPA system gains traction with legitimate publishers, the scam publishers may start to be isolated in the CPC system.  If this happens, Google’s AdSense model may reach a tipping point where CPC rates start to drop because the quality publishers are no longer subsidizing the scammers.  This will help make the Internet a better place (To this point I disagree with Scott Karp’s conclusion that Google’s CPA network will create direct response hell online.  It is easy to dupe users into clicking on things, you have to deliver value to them to get them to sign up for a service, order a product, etc.)   

Consumers Still Aren’t Getting the Full Value of their Attention with Traditional CPA

• This is why Google’s news ultimately lands with a thud.  If we assume Google is wildly successful with their CPA model, it may help elimiate the search engine spam and click optimized sites that frustrate untold online consumers.  This is a good thing for consumers but it doesn’t go nearly far enough to deliver consumers the true value of their attention online (see my post on Buying Attention here).  At Jellyfish we take Cost Per Action one step further to create Value Per Action, which directly rewards consumers for making purchase decisions on our site.  In the Jellyfish VPA model, as advertisers bid higher and higher rates to reach buying consumers, the end consumers share directly in that competition for their attention through greater savings.  This won’t happen in Google’s CPA system.  As advertisers compete for CPA opportunities and bid up CPA rates, Google and third party publishers will make more money, but the end consumer will still be completely disconnected from this significant value creation.  Ultimately, this makes Google’s CPA moves of marginal benefit to online consumers.           

Since our launch in June of 2006, Jellyfish has been working hard to drive adoption of a new form of online advertising we call Value Per Action that eliminates the waste from traditional advertising and transforms it into added customer savings (you can read about VPA here and our vision of advertising here). 

With Jellyfish taking the high road against traditional forms of biased, interruptive advertising, many questioned how we were going to attract customers and generate awareness for our brand and our shopping search engine.  Not a bad question and one we asked ourselves many times.  Television ads, billboards, and even Google keywords were out of the question.  But what was left? 

The answer for Jellyfish has been something we call Smack Shopping.  The Internet’s first live shopping game show, Smack Shopping combines the fun of a game show with the fantastic deals generated by the Jellyfish cash back auction.  Smack shows run every business day at 12 CST and include multiple Smack Auctions where a limited quantity of products are offered for sale at increasing cash discounts.  Participants have the chance to buy the hottest online products at remarkable discounts, play games for prizes and interact with a virtual community of other Smack shoppers (a Smack by definition is a group of Jellyfish).

To our knowledge, Smack Shopping is the first game show broadcast live to an online audience that actively participates in the event.  There are other companies working on streaming video plus chat, see for example Pete Cashmore’s discussion of Lycos Cinema here and the much anticipated launch of Joost here, but Smack Shopping does more than view pre-assembled video.           

Fueled by Jellyfish cash back, the game and the entertainment is actually advertising, but I doubt any of our users would view it as such.  It’s a customer acquisition strategy that is consistent with the Jellyfish promise.  Instead of spending money interrupting people with traditional advertising, Jellyfish is using advertising dollars to create compelling game show like content that consumers seek out and invite into their lives.  And it is working.  From the beta launch of Smack Shopping on November 1 until today, Jellyfish has grown its user base over 5X, with Smack Shows regularly engaging 10’s of thousands of online shoppers for an hour or more of fun every day.   

Advertising as content.  It is a trend we are going to see more of as consumers gain more control over where they devote their attention and advertising seeks out new ways to be relevant.  We view Smack Shopping as permission-based, engaging advertising at its best. 

Brian Smith’s post at ComparisonEngines today has some very interesting news regarding Shopping.com’s upcoming test of a Cost Per Action advertising model at its popular comparison shopping site.

The move by Shopping.com is a small step in what I see as the inevitable march towards Cost Per Action advertising. 

But what makes this even more interesting is the way in which Shopping.com is making the transition from clicks to sales conversions.  Taking a page out of the Google Checkout playbook, Shopping.com is launching their CPA experiment within a Universal Shopping Cart in which Shopping will take control of the entire order process AND a large part of the end customer relationship.  Essentially, this universal cart would allow Shopping.com to become a new type of super retailer-a Shopping.com customer can add products from any merchant in the search engine into a single Shopping.com cart and purchase the products directly through Shopping.com without ever having to go to the retailer’s site.

Why would Shopping.com transition to CPA advertising in this way?  In addition to the benefits of CPA advertising in general, I see three specific benefits:

1. They solve the Pay Per Click search engine loyalty problem.  Currently, Shopping.com and other PPC engines have a difficult time creating loyalty with their existing customers because they lack a means to create a strong ongoing relationship.  By setting up a Universal Cart, Shopping.com can create a new user-friendly feature and establish a stronger, ongoing relationship with each customer.
2. They learn more about each customer.  Customers using the Universal Cart will not only store their own personal data with Shopping.com, they will also share their actual purchasing habits directly with the engine. 
3. They enable Cost Per Sale advertising without having to track sales at third party merchant sites.  This has been a big hurdle to the adoption of CPA advertising; intermediaries have been leery of relying on third party stores to track and report their conversions at the merchant point of sale.  Like Google Checkout, Shopping.com has sought to avoid this problem by taking control of the sale itself.  This provides them with a direct and powerful new way to record the sales they generate and in turn, maintain direct control of their revenue. 

But will retailers go for this?  Will they turn over their customer relationship to Shopping.com in this way in exchange for the accountability of Cost Per Sale advertising?  Brian Smith raises these concerns in his post, and many others have already discussed the push back Google Checkout has gotten for controlling the customer (see Greg Sterling’s post here for example).  It will be fascinating to watch this play out.  Frankly, I will be amazed to see a large number of retailers essentially agree to give up a big part of the customer ownership equation and transition into more of a product shipping/wholesaler entity and away from being a customer-centric retailer. 

There is an alternative way to set up CPA advertising that avoids the customer ownership dilemma being pushed by Google Checkout and now Shopping.com.  Namely, allow the end retailer to continue to control the order process and customer relationship and set up a mechanism to record converted sales at the merchant point of sale rather than taking over the sale itself.  This traditional method has been used for years in standard affiliate marketing and it is the method we use at Jellyfish.  I assume you can guess which method I think is better.  Time (and customer demand) will certainly tell whether retailers decide they are willing to cede ownership of the order/customer in exchange for the sale. 

Mark Boslet at Dow Jones wrote a great piece on this subject yesterday “Internet Cos, Advertisers Study Changing Online Ad Market.”

The first sentence of the article makes it clear the potential game changing nature of CPA:

 “A potential shift in how online advertising is sold and billed has advertisers and Internet companies scrambling to understand its full effects.”

My take away from the article is this: Pay Per Click advertising is too profitable for CPA to take over at present.  Sure CPA eliminates risk for advertisers, but why should intermediaries like Google, Yahoo! and MSN take more risk if they don’t have too?   

The current pay per click system may hang on as the dominant model for several years, but CPA is inevitable in my opinion.  Technology now makes it possible to deliver advertisers 100% accountability and certainty in their advertising.  This is an extremely powerful concept that eliminates a tremendous amount of waste from advertising; waste that could and should be reallocated to end consumers. 

At the end of the day, advertisers are in the business of selling products & services, not clicks.  The transition from the Gross National Click Economy online is coming and I hope Jellyfish and its new VPA advertising model leads the charge and moves this transition forward.