Google's Foray Into Cost Per Action (CPA)

Enough people have asked me about Google's beta test in CPA marketing that I guess I have to write something about it, so here goes.

First, a little background - it was announced last week that Google is testing cost-per-action distribution on part of their network. In other words, instead of paying a publisher for every click on their site, they instead paid them on actual conversions that result from these ads (i.e., every person that fills out a mortgage form from an ad on a publisher site would result in a $10 flat fee to the publisher, irregardless of how many clicks it took to get that one conversion).

Upon hearing the news that Google was dipping its feet into the CPA waters, many affiliate marketing and lead generation companies (like Commission Junction, Linkshare, Adteractive, Azoogle, or Quinstreet) etc must have been quite worried (it reminded me of a great Onion article - "Dolphins Evolve Opposable Thumbs; 'Oh, Shit,' Says Humanity"). After all, when Google enters your market - especially if it online advertising-related - you've got to be a little worried.

I can imagine a scenario where a Google CPA network could result in the destruction of the CPA industry as we know it. With a team of dozens of engineers developing cool algorithms to optimize CPA conversions, an existing massive distribution network, and Google's good name, it would be fairly easy for Google to grab a large chunk of the CPA market. Heck, Google could even combine AdSense with a CPA network and test the two products interchangeably until they have a clear winner from a monetization perspective.
Put yourself in the shoes of an online publisher that currently relies on affiliate marketing and AdSense for revenue. Given the choice between guessing which offers from Commission Junction will work for your business and letting Google optimize the entire process for you (assumedly with superior results), this seems like a slam dunk for any publisher.

Frankly, I've felt for years now that - when it comes to online marketing- Google has the resources necessary to conquer whatever vertical or distribution-type they want to. And CPA seems to fall into that bucket.

The real question, then, is not whether Google could bite into CPA, but whether it is actually beneficial for them to do so. After a few pensive walks around my 1000 square foot condo (short walks!), right now I believe the cons outweigh the pros. Here's why.

First, let's start with the advantages of a CPA network:

1. Even More Market Share: So Google already owns the CPC market, and with increasing competition from MSN, Yahoo, and Ask, I doubt there is a lot more room for additional market share. It makes sense, then, to try to dominate other marketing channels. Google has already demonstrated an interest in this with the acquisition of dMarc (radio ads), the Google print beta, Google site inclusion (CPM), and the new Google Video Ads. A CPC network is yet another extension of Google's overall strategy - use CPC as a base and work on expanding into other marketing channels.
2. It Benefits Google's CPC Product: As noted earlier, if you can create a hybrid distribution network that combines CPC and CPA, you are likely to end up being able to offer higher payouts than a CPC or CPA only network (assuming you can optimize the results accordingly). This means that the existing AdSense network could be even stronger with a CPA component.
3. Every Google Wants to Rule the World: As Jay Weintraub speculates in his blog, a CPA product may be part of Google's overall strategy to combine its "GPay" online payment system with online advertising services. In other words, end-to-end ownership of the entire transaction between consumer and merchant (but see below: this smells of anti-trust issues . . .).
4. Capitalizing on a Trend: There's no question that CPA marketing is hot right now (recall that I predicted in my 2006 predictions post that this would be the "year of CPA"). Google may think it needs to start developing a CPA product to be ready when CPA really takes off.
5. A Hedge Against Click Fraud: Jay Weintraub also gets credit for this one - it is much more difficult to "game" CPA offers as compared to gaming CPC offers (through pretty simply click fraud techniques). Google is under increasing pressure to address click fraud and perhaps a transition to CPA will reduce the click fraud heat over time.
6. More Control of AdWords/AdSense: If a world someday existed where Google was the dominate affiliate marketing distribution network (replacing Commission Junction and Linkshare), this could give Google a lot more leverage over the dreaded "arbitrage" affiliates that create multiple stealth Web sites to promote online casinos, or credit card offers, and so on. If Google was the biggest CPA network, it would be much easier to restrict affiliates from gaming the AdWords/AdSense system. In other words: if you want to participate in the Google CPA network, you have to police your affiliates on AdWords. A very nice carrot and stick in my opinion!

OK, OK, that's a lot of very plausible reasons for Google to create a full-on CPA network. Now let's look at some of the downsides:

1. It Ain't As Easy As It Looks: Replicating Commission Junction is probably not that difficult - it's a pretty basic user interface with limited technology (at least by Google's standards) and a critical mass of customers that already probably do business with Google anyway. Replicating a LowerMyBills for mortgage offers or a Quinstreet for education offers, on the other hand, takes a lot of work. For Google to achieve the eCPMs that these folks have achieved after years of focus, wrong-turns, development of expertise, and aggregation of business relationships, is going to be difficult. An algorithm can't call 2000 mortgage brokers and create a high-paying mortgage offer. And Google wants to do everything with computers so it seems unlikely that they could really compete against the niche players in CPA (niche in focus only, not in revenue and profit).
2. CPA May Cut Into Google Profits: Right now, there are a lot of dumb advertisers buying CPC ads on Google. By dumb, I mean that there are advertisers that don't track conversions, engage in vanity-based bidding battles for top position (irregardless of ROI), and in general have no idea whether Google is really making them money or not. For these advertisers, Google really has no incentive to eliminate click fraud or create a product that provides a greater likelihood of positive ROI. My sense is that Google actually makes a lot of money from ignorant advertisers - switching to a model that increases profit transparency could actually have a negative impact on Google's revenue.
3. CPA is a Slippery Slope: Give someone an inch and they'll take a mile. If Google starts offering some placements on a CPA-basis, I have no doubt that advertisers will slowly start to ask for more and more CPA and less and less CPC. And why stop at CPA placements? Savvy advertisers will begin to push Google for revenue share deals as well (as is essentially happening in the CPA industry at the moment). As you go closer to rev-share and farther away from CPC, you increase risk for the publisher (Google) while decreasing risk for the advertiser. That's something that Google probably wants to avoid.
4. Do What You Do Well: As noted, Google is seeing a lot of competition these days in the CPC space, in particular from the new and impressive MSN AdCenter, the upcoming release of "Project Panama" from Yahoo, and Ask's major marketing push. I think Google would be better off focusing on making AdWords as good as it can be at the moment, rather than going off in a million different monetization directions at once.
5. Too Much Power in Google's Hands?: I've mentioned this before - I really believe Google has become a big enough force in a big enough industry that they have to start being carefully about wantonly taking over new industries, simply because too much power may eventually put them under the scrutiny of the DOJ's anti-trust unit.

In the end, it's way too early to tell what Google plans to do with this CPA test. I do believe that CPA is something that all online distribution networks are going to have to confront in the coming years, especially if the market tightens up again like it did after the dot com bubble burst around 2001. Right now, though, Google is on top of the world, thanks almost entirely to its CPC products. With a lot of risks associated with any entree into CPA, my bet is that this is nothing more than a fun beta that won't see any legs for some time to come!

Google Video Ads - Ahead of Its Time, Or At Least Its Advertisers

Today I got an invitation to attend a training session for Google video ads. Internet video is the latest "new, new thing" to sweep through Silicon Valley - not unlike mobile commerce, local search, procurement, etc, etc. For about six months, you'll see dozens of start-ups getting funding, ridiculous projections of market size from inane news sources like ClickZ, and probably two or three conferences dedicated to Internet video. So it only makes sense that Google jumps on the bandwagon and prepares a product around video, right?

Well, yes and no. The problem is this: Internet video is exactly the kind of product that Google's advertiser won't use, simply because it is too technical, requires too much effort, and won't be sold on a CPC basis. But that doesn't mean it won't eventually be a good idea. Read on.

Let's start with Google's advertisers. Last I heard, Google had over 600,000 of them. The vast majority are "mom and pop" companies like a local dentist or a mail order helmet company. The big players tend to be "direct marketing" companies like eBay, Nextag, or Amazon.

Obviously, small advertisers aren't the target audience for this product. The beauty of Google AdWords is that it enables small advertiser with no marketing resources or graphic design skills to effectively compete in online advertising. All you need to do is buy a few keywords, create a text ad, send people to a reasonably good Web site and voila, you're an online marketing expert!
Compare that process to hiring a video production company to film your video ad, finding a voice-over artist, and generally paying tens of thousands of dollars just to produce your video (prior to getting any traffic whatsoever) and its pretty clear that Google Video Ads is the polar opposite of what the majority of Google's advertisers need or could possibly use.

So, then, I guess Google Video is intended for the eBays and Amazons of the world? Well, maybe. It's true that eBay is spending a lot of money on TV ads these days, and there is certainly a lot of TV and radio advertising by travel companies like Orbitz or Travelocity.

I think, however, that Google is going to have an uphill battle getting these companies to invest a lot of resources in Video Ads. First, these big players are slow-movers. They generally have bloated bureaucracies and new marketing products get lost in the shuffle. Moreover, Google is in their "CPC" budget not their "branding" budget, which means that either the CPC team is going to have to buy branding ads (which will upset the branding team), or the branding team will have to buy branding ads on Google (which will definitely upset the CPC team!).

A comparable product failure Google just endured (you'd think that they'd learn . . .) is Google Print Ads - the 'auction' for space in Ziff Davis magazines. Buying print ads through Google doesn't appeal to the small guys (who don't have the resources or knowledge) and the big players will just go directly through the magazines or their agencies.

In sum, then, Google Video is basically a product without existing customers. So why is Google launching this product?

Well, Google has not made a secret of its desire to capture the "G1000" - the top 1000 advertisers in the US. These are the Fords, McDonalds, and Mastercards of the world, the companies that love clever ads from clever advertising agencies. Google Video - along with Google Print, Google Radio, and Google Site Targeting (CPM) - is the carrot to get these big (and dare I say, dumb) advertisers into the fold.

CPC advertising is about as familiar to General Motors as branding is to an affiliate marketer. So if a Google ad rep walks into GM's ad agency's cushy offices and proposes a 10 million click campaign, they'll get a lot of dumb looks. But start talking about "video" and "lift" and "branding" and suddenly the ad agency wakes up! Video is an opportunity to be creative, to be clever, and best of all, to measure success based on "lift" and not on "ROI."

GM isn't going to drive up Google's stock price by purchasing millions of dollars of video ads (or site targeting). But hey, maybe you get GM in the door with video, and then once you've established a rapport with the marketing folks, you slowly introduce some CPC ads into the mix. And maybe a little later on, you start showing some metrics to GM - about how CPC is far more impactful on the bottom line than those one million dollar Super Bowl ads the ad agency keeps pushing.

Suddenly, GM realizes that those Google Video Ads are pretty silly, and so is the hundreds of millions of dollars they're spending on ad agency and big TV ads. One GM can spend as much on CPC as 10,000 small advertisers, and it's a lot easier to manage one company versus 10,000.

So will Google Video take off in the next six months? Of course not. But is it an entree into a pool of new clients Google has been desperately seeking for over year? Possibly, we'll have to wait and see.

How Much is $60 Million Worth? About 25 Cents . . .

Right now I'm sitting at home watching game five of the NBA finals. Mind you, it's not because I like the NBA or have watched any of the prior playoff games, I'm just bored and there's no World Cup games on at the moment (other than the replays on the Spanish channels).

According to Ad Age, the marketing story of the NBA Finals is the fact that both the Dallas and Miami arenas are sponsored by American Airlines. As the story notes: "American Airlines could reap more than $60 million in brand exposure as a result of the National Basketball Association Finals between the Miami Heat and Dallas Mavericks that begin tomorrow night." This is apparently quite a windfall, because American Airlines "only" paid something like $20 million for the naming rights for these areas over the course of something like 10 years.

So $60 million minus $20 million - that's a $40 million profit, and 200% margins. Surely some vice president of branding is getting upgraded to a corner office at American Airlines' headquarters.

Ah, if only it was that easy. Let's think about this for a minute. American Airlines sells air travel. Most people I know base their air travel on three factors: 1) price; 2) schedule; 3) frequent flyer membership. In other words, in the event of equal price and schedule, people will opt for the airline where they can get the most frequent flyer benefits, but for the most part, air travel is a commodity, like table salt, electricity, or paper napkins.

That begs the question: what exactly is the "benefit" American Airlines receives from all of this branding? Will travelers opt to pay higher fares to fly on American Airlines (instead of a competitor) as a result of seeing their name emblazened on NBA Arenas? Will viewers flock to American Airlines' Web site to learn more about the company products and offerings? Will frequent travelers switch to the American Airlines frequent flyer program as a way of supporting their favorite NBA team?

Of course not. In short, I doubt that American Airlines will reap any financial benefit from the increased exposure from the NBA Finals. I suppose branding works when you are marketing a car, or perfume, or something that people make an emotional connection to prior to a purchase. But no one cares what airline they use. Assuming the seats are about the same width, the safety record is fine, the schedule works, and the price is right, it just doesn't matter.

So let's revisit that $60 million of branding exposure. If that's how much it would cost to buy advertising during the finals, I guess that's a good deal if you really wanted to buy that much advertising for your commodity-product (can you imagine Morton's Salt paying $60 million for branding? I can't). But you aren't going to see an additional $60 million of profit or even revenue on the American Airlines quarterly earnings report.

Which would you rather have: a) exclusive branding rights to two arenas, or b)120,000,000 clicks on Google and Yahoo? If you answer "a", I've got a bridge to sell you in New York . . .