advertisement
advertisement
advertisement

Search Advertising And Auction Transparency

What do the Federal Reserve’s Quantitative Easing program (which ends this month) and online paid search have in common? More than you might think–as pricing for both use a form of auction models.

What do the Federal Reserve’s Quantitative Easing program (which ends this month) and online paid search have in common? More than you might think–as pricing for both use a form of auction models.

advertisement
advertisement

The increasing use of auction models to purchase online advertising has significantly improved the efficiency and effectiveness of the online channel. Executed properly, auction models align supply and demand in near real time and help buyers understand the true price elasticity of the product or service being purchased. This type of transparency is long overdue in the world of media.

The largest online auction marketplace is for paid search, which Google dominates. Google controls more than 70 % of the $12B annual paid-search market in the United States. There is much to admire about Google both as a company and for the products it creates. Being responsible for spending my clients’ advertising dollars wisely, I am eternally grateful for the large auction model that Google has made ubiquitous for online advertising. But is Google’s bidding model for paid search as transparent as it could be? To answer this question, let’s look at the history of the auction model for paid search.

Google did not create the search-advertising auction business. A little company called Goto built the first auction marketplace for search in 1997. On Goto, advertisers could submit a bid for a particular key word on a per-click basis, not per impression as had been done up to that point. We could now buy “actions” instead of “eyeballs.” Advertisers’ page rankings were based on the descending order of their bids. This model was initially very successful, but it suffered from volatile pricing because faster bidding technology enabled bid manipulation. It’s worth noting that, as advertisers, we could see each other’s bids, a revolutionary level of transparency in media buying.

Overture

Years later, Goto became Overture and the company was eventually purchased by Yahoo. Yahoo wanted part of the auction marketplace since its significant traffic helped power Google’s rise to become the dominant search engine. Google supplied natural search results for Yahoo for several years.
Google sold advertising on a fixed-cost basis for the top two search results on its engine but eventually began to experiment with a cost-per-click pricing model that was used by its competitor Overture (previously Goto). Google made improvements to the auction model to mitigate the pricing manipulation present in Goto’s model.

Additionally, Google added a quality element to the equation for page rank. The quality element was a score Google gave each website for the keyword being searched. This customer-centric approach delivered a much better use experience. Advertisers could not simply buy their way into the top position; they had to deliver content that was relevant to the user’s search. This was a brilliant modification because it helped consumers find what they were looking for, it made content creators better at delivering relevance, and it helped keep advertisers honest. This approach helped Google grow into the powerhouse it is today.

advertisement

By using quality and bid price to determine a paid ad’s position in search results, Google created a new way of ranking results. The ultimate rank for any paid search keyword is a combination of quality score and bid. However, unlike the original offering by Goto, bids can’t be seen by other agents; each advertiser only knows its own bids. Google eventually released a measure of their quality score that is only visible to advertisers and not to the public or competitors. The addition of a quality score was immensely helpful in driving more relevant content creation. The costs of a poorly executed website and its ensuing inability to attract its target audience were immediately clear.

Eventually Yahoo decided to change the Yahoo Search auction to mirror that of Google’s, so that we could no longer see the active bidding from all agents. No doubt a large part of this decision was driven by the fact that the revenue-per-search term for Google’s auction was much higher than Yahoo’s. While advertisers were still participating in an auction, it was not as transparent as before. It is also worth noting that Bing, the second most popular search engine but significantly behind Google in terms of market share, also operates in a similar fashion.

For economics enthusiasts, Google’s approach is technically referred to as a Generalized Second Price Auction model, and it differs from an English Auction model where the price from all bidders is known. More detail can be found on these Wikipedia pages.

But what if we were to increase the level of auction transparency and add elements of the English Auction model while maintaining all the positive elements of Google’s current model? What if we could see the actual bids as well as the quality scores for each advertiser after the bid is accepted? Along with freezing changes to the bid for a period of time, post-acceptance bid exposure would prevent gaming during the bidding process.

This suggestion could be construed more with what’s been coined as a Generalized English Auction. Here, an auction clock drives the bid price. Advertisers in the auction can see the current price continuously increase over time and place their bids by dropping out once the clock displays their reservation price. The clock stops once the next to last advertiser drops out. In this type of auction, bidders are ignorant of individual bids, but do have historical knowledge of prices at which advertisers have dropped out from previous auctions. A detailed review of this concept can be found in the American Economic Review 97(1), 2007 pp 242-259.

I recognize that there are layers of complexity here better left to PhD economists. However, I also believe there are some obvious business benefits worth contemplating. For starters, advertisers would understand how a much higher quality score can produce lower prices. The beauty of Google’s system is that it rewards highly competitive ads and relevant content with lower prices–creating incentives for good behavior and punishing the poor ads with higher prices. This additional layer of transparency would lead to better content and more relevant ads.

advertisement

Imagine an ad agency showing a client a screen capture where its biggest competitor is paying $1 less at auction for their most trafficked keyword because of a 9-10 quality score. This discrepancy would motivate the client to improve content and relevance for high-volume key words.

This approach would likely also decrease short-term revenue for Google because advertisers would estimate the appropriate bid ceiling for any keyword. But it would also force advertisers to become more relevant more quickly. With better performing ads, marketers are likely to spend more in digital media. In the long run, better ads would make both Google and advertisers more money.

Transparency aligns the interests of advertisers and Google to make the most out of the marketplace. Google has been a leader in helping to bring a much needed, large-scale auction model to online advertising. I am hopeful that they will continue to innovate to better serve end users, advertisers and shareholders.

advertisement
advertisement

About the author

Steve has over 24 years of agency and client side experience leading CRM, interactive marketing, sales and media practices for brands including Nissan, Bank of America, Visa and Procter & Gamble, to name a few. In 2011, he was named an Adweek Media-All Star for his innovative work measuring earned and owned media content and developing predictive analytics models to optimize digital ecosystems

More