Google has entered an arena which comes as no surprise but still a slight shock. The company has made billions of dollars off click based advertising, literally billions. In fact, they make more than a billion dollars each month from advertisers paying pennies per click to tens of dollars. Google’s methodology for generating revenue, though, runs counter to the majority of their advertisers’. Google makes money, billions (did we say that already?) from users clicking on links. Except for a relatively small number of advertisers, i.e. the pure click based arbitragers (see our articles on PPC vs. Affiliate Arbitrage – Part 1 and Part 2) which to some degree includes comparison engines, the rest of the advertisers don’t. Many of Google’s advertisers still make their money through some form of arbitrage, encompassing the traditional mortgage lead generation sites to the less common but more recent trend among newspapers to buy clicks and earn money from their CPM advertisers.
When it comes to arbitrage, we can think of it in the aforementioned PPC vs. Affiliate as well as direct response vs. branding. We can also think of it not in the type of arbitrage, i.e. clicks to clicks or clicks to actions or clicks to impressions, but the duration – short versus long. In all of the types mentioned thus far, they fall into the short category. An advertiser who buys a click on Google will generally know not in days but hours or minutes whether that click made them any money. Some users might convert the second or third time after clicking on the same ad, but the vast majority of converters do so then. The long scenario works in reverse. A small percent might convert right away, but the vast majority who do, will do so later. The difference usually comes down to the complexity of the transaction. Think of eBay or Circuit City. For some items – when someone finds exactly what they want and/or for lower priced goods, etc., they might purchase that second, but for more transactions than not, users visiting the site fall along a spectrum that is much less binary than the yes/no of the short form of arbitrage. The long form still buys traffic to make money later; they just have to do more work/science/guesswork to do so profitably.
The long form of arbitrage includes some incredibly well known brands and large spenders. The short form of arbitrage does as well, but it has more spenders than big brands. Outside of duration of conversion, the real distinction comes down to the likelihood of being able to assign a value to a transaction. Score one here for the short arbitragers, especially those in some form of lead generation. They spend hundreds of millions monthly trying to find the clicks that will lead to profitable actions. In the process, they waste a lot of money which leads to Google making more money. Some like it this way – the smart ones can use technology along with marketing savvy (sometimes just the latter) to make a decent margin on the spread between clicks and actions. The majority that we know, though, continue doing this but at a lower margin than before. Competition along with Google’s being Google have made it more challenging. Those that do not have the sophistication have it even worse. Overwhelmed by the complexity of a relatively easy concept – bidding on keywords – many won’t reach their potential.
What could potentially erase the opportunities the arbitragers leverage while increasing the competitiveness of those without the same level of paid search sophistication? Advertisers being able to buy search traffic on a CPA basis. From the perspective of an arbitrager, Google has threatened to do this for some time, beginning in June of 2006 when they announced their first CPA venture, the content referral network, which they followed up with Pay per action beta in March of this year with the official launch taking place towards the end of July. With their usual genius, Google leveraged their expansive publisher network to gain the necessary insight into performance – leveraging their long tail of often underwhelming publisher placements – as well as building up an increasing stable of advertisers. With their Pay per action program, Google didn’t just create an interface for publishers, they announced that they would start to rotate in CPA ads into certain publishers’ placements. Having an interface for publishers covers one piece of the landscape, but the larger opportunity comes from being able to choose when and where to show an ad for maximum yield. Having all those pieces in place – CPA advertisers, the publisher placements and acceptance, Google could then connect the two and offer advertisers a true CPA product. And, that’s just what they did on the 24th of September.
Given the lead-up, that is why we call their announcement – the Conversion Optimizer beta, “a cost-per-acquisition (CPA) bidding tool” a surprise but not a shock. It’s a surprise because it seems to run counter to their core business of clicks and against the “if it ain’t broke don’t fix it school of thought.” Google has trained upwards of a million advertisers to buy ad space without interacting with a person and to do so on a cost per click basis. If it truly worked for the advertiser didn’t concern them – only search marketshare and making sure the traffic met certain standards so that the advertisers had a chance to succeed. That Google will now offer CPA optimization for advertisers comes as no shock for more Google-like reasons. They have a history of trying to take out the middleman, of instituting non-affiliate friendly policies. Affiliates act as a large outsourced CPA bid-management system for advertisers. Google has the insight, the technology, and the upside by taking an affiliate’s margin in-house. Disintermediation is but one reason. Growth is the more likely culprit. Google always knew they would have to enter this game. For them it wasn’t a quest of why but when. They might continue to see market share creep but, slowly, the same holds true with CPC prices, with their latest non-market driven approach, Quality score, having helped.
Google does things for Google. This move might seem to help the advertisers, and indeed it will for many, but the desire to help them didn’t drive the change. The need to make more money did. Unlike a true arbitrager, Google doesn’t take on any real risk. They do not pay for inventory on a fixed cost basis where they must rely on optimization to make them profitable. They will have an opportunity cost as they test, but if anyone has the mathematical horsepower to make it work, they do. As with everything Google, nothing comes in black and white, just shades of gray. Both disintermediation and growth have some truth to them. Regarding the former, it won’t stop arbitrage. Similarly, many advertisers will try out this CPA approach only to realize they do not receive the level of traffic they want. It will force them to make changes – to use other Google products like website optimizer, to raise their CPA, or equally helpful to Google, to go back to CPC and raise bids. In the end, CPA Optimizer moves Google one step closer to becoming a true broadcast company and take on the run of network display business in advance of their integration of Doubleclick, or perhaps in spite of Doubleclick.
Network
