POV: Finding The Optimal Position For Your Ads

Written By Andrey Milyan | June 24, 2011 | No Comments
Optimal Position

Photo courtesy of InternetMarketingStop.com

Google AdWords and Microsoft adCenter both offer a vibrant, real-time, auction-based environment for their advertisers. At the same time, these platforms put a lot of pressure on an advertiser to “win” an action by outbidding the competition and appearing at the top of the search results…

Situation
Google AdWords and Microsoft adCenter both offer a vibrant, real-time, auction-based environment for their advertisers. At the same time, these platforms put a lot of pressure on an advertiser to “win” an action by outbidding the competition and appearing at the top of the search results. Such a strategy, while capturing the maximum volume, does not necessary result in the most optimal outcome for an advertiser. To test this hypothesis, Reprise used a branded head term for a well-known, national brand. There were several reasons why this particular keyword was selected:

  • It is a very high volume branded term that converts well but the intent behind the query is somewhat ambiguous since it does not refer to any specific product offered by the company.
  • The keyword faces a lot of competition from resellers and affiliates, resulting in a dynamic and often unstable market.

In a competitive auction environment like the one for the test keyword, all advertisers set fairly high bids to prevent negative impact on their campaigns from sudden changes in the bidding behavior of the competitors. Because of the smart bidding features, an advertiser does not pay the full bid amount. Instead, a winning bid/quality score combination ad pays only a fraction higher CPC than an ad below it, assuming the quality score for both ads is roughly the same.
For example, if advertiser A is paying $0.60 CPC for the test term in a given market, resulting in an average position of 1.5, advertiser B could be paying $0.61 for an average position 1. The actual bids could be as high as $5 or $10. Remember, the bid is an indicator of the ceiling cost per click a given advertiser is willing to pay. In a stable environment, the CPC remains low. However, whenever an advertiser increases its bid to improve the average position, the algorithm recalculates the average positions based on the new bid amounts and quality scores. Since the bids are set high, the CPC begins to rise, often times without any actual change in the order of the ads on the page. Once that particular advertiser realizes that the competitors’ bids are set too high and stops raising its own bid, the auction market stabilizes but now at a much higher price. Advertiser B could now be paying $2.01 per click for position 1 and advertiser A could end up paying $2 per click for position 1.5 and so on.
Microsoft adCenter Example This is precisely what happened with the test keyword on adCenter. With a number of new advertisers entering the market (mostly affiliates and resellers), the actual CPC started rising over a period of a few weeks. At a certain point, it became clear that even position 1.5 (which Reprise found to be historically efficient for this advertiser) was no longer cost-effective. To test the hypothesis, the ads were pulled back to an average position of 2 by dramatically lowering the actual bid across all markets. The results were striking:
adCenter

  • Just by lowering bids and moving closer to position 2, Reprise was able to lower the media cost for the test keyword by 88.1%. Likewise, the cost per click went down 85.3%.
  • While the clicks did go down as a result of the lower CTR (which itself is a result of the lower average position), the percentage drop was much lower than for spend, at only 19.2%.
  • Even more interestingly, the absolute number of actions went down just 3.4% while the cost per action went down by 87.7%.
  • As a result of the lower average position, the CTR did go down 27.6%, resulting in a loss of volume. However, this was a small price to pay for a greatly improved efficiency.

Conclusion While moving to a lower position is not always desirable or prudent, the adCenter example above illustrates the importance of paying attention to the efficiency of a keyword at a given average position. While some battles are fought and won with a brute force of a high bid, it is important to pick your battles. Given the available level of control, Reprise continually monitors the top volume keywords for any efficiency that can be extracted from adjusting the average position of an ad. The budgets are constantly adjusted to make sure the most efficient ads and keywords get the maximum exposure.

Tags:

Leave a Reply