Somewhere between the time when paid search first got recognized as a marketing tool and, well, last week, advertisers asked in vain about the importance of being in the first position in search. Of course, since the dawn of processed wood fiber directories, advertisers have been duking it out for positions in phone books, but those advertisers did not (and still do not) have the Atlas Institute throwing intelligence their way.
“Be number one, stay ahead of your competition!” goes the mantra. But how important is it to be number one? For a short while, paid search listings were only the top three or so listings syndicated from the likes of Overture and Google. Not being in the top three listings meant enormous traffic drop-offs and equally enormous cost disparities if your listing fell out of syndication.
That’s not quite the case anymore. Google is a search destination (much in the same way Weather.com is a vacation destination) and the right rail is filled with paid listings. Overture isn’t a search destination, but Overture paid listings on Yahoo! and (for now) MSN contain far more than the first three listings for keywords. Phone books have been around since 1880, but the impact of performance and cost disparities for position one through ten is far more important today than it was even a year ago.
Think first to list
The thought process behind "The Atlas Rank Report, Part One: How Search Engine Rank Impacts Traffic" as well as "Part Two: How Search Engine Rank Impacts Conversions" appears to be quite simple. Empirical evidence supports that a top position achieves higher traffic than a lower position, but the important question to ask is this: how much revenue is lost to lower positions?
The answer, my friends, is blowing through your paid search budget. Atlas applied two interesting dynamics to measuring search effectiveness in part one of the report: “Relative CTR (Click Through Rate)” and “relative impressions,” both of which use the first position as a benchmark.
In a perfect world, everything uses the first anything as a benchmark to activity. Every bit of data presented depends on a theory that the first position generates the highest number of impressions and clicks. If that theory is not valid, then you can throw out the following.
The Atlas reports concede that titles and descriptions effect rank but these measures claim to be made independent of any effect titles and descriptions might have. Huh? This means titles and descriptions are important, but traffic gets dropped by position so often that patterns can be drawn, even in the presence of changes in a listing’s appearance and the keyword’s relevance.
Ok, so maybe throwing out all of this is a bit harsh. Let’s say, you have accept the benchmark to buy into the results of the report. A second calculation in part one assesses the potential traffic to be gained by using the product of relative impressions and relative CTR, deemed “click potential.”
If you retain and apply nothing else from the first report, remember the calculations for assessing click potential. In addition to providing a good generic benchmark for traffic loss, as the report notes, it dispels the myth that advertisers must be in the top three results. And while I’d say the top-three myth was practical lore based on historical listing syndication requirements, the click potential calculation might serve as an effective means of estimating a future search budget.
In the example listed in part one, using an average CTR for position one of 7 percent, a term listed at position one on Overture will generate 70 clicks (100 percent click potential) if a thousand searches are performed within a given time period. However, position three (often considerably cheaper) will only generate 41 clicks (58.8 percent of click potential.) Now, apply a vendor provided or historical cost dynamic (average cost per click) and you will have an educated guess as to future keyword costs based on response drop-off rates for positions beyond the top three. Just remember that Google differs from Overture in its relationship with click costs and positions.
Achieving desired action euphoria
Part two reports on how positioning might affect conversion behavior through analysis of over 41 million clicks in July and August, 2004. That is, conversions from keywords were monitored as they moved down the ranks of search results. Again, in order to keep the results from being biased by a category or specific industry, the report relies upon response activity relative to position one.
“Relative conversion rate” is a measure of the expected change in conversion compared to the first position. “Conversion potential” is a measure of an anticipated change in conversions in relation to position one.
In the report’s example, if the relative conversion rate is 10 percent for position one on Google, position two will generate only a 9.1 percent relative conversion rate. Though position two generates a high conversion rate compared to position one, the second position receives only 58.9 percent of click volume. When you multiply the lower conversion rate by the click volume, the conversion potential is illustrated in 54.5 percent fewer conversions.
The report shows as positions fall, so do conversion rates. And, as was the case for the first report, this model works as an effective means to calculate projected costs. The exception in this instance is that you can also apply the response or conversion cost dynamic to projections. The true value of this data is in illustrating that even though position two might be less expensive from the click cost perspective when you add up conversion costs, a lower position does not mean a lower cost per acquisition.
Get high, but don’t forget to go low
Weird things always happen if you toss data or ideas around long enough. It reminds me of the time I spent five days (that is, I'm pretty sure they were days) at a tea room in Prague sucking on a hookah while reading a translation of Marcus Aurelius’ meditations. Upon reading the great emperor’s notes, I was able to surmise that every self-help or professional development book I ever read had stolen its foundation from him. For some reason, the only word I could muster was an angry, “dude!” Black was white, night was day, nothing made sense anymore and what the heck was in that hookah?
The Atlas report uncovers a search anomaly equally as devastating, and they were able to come up with something better than a one word generic expletive: a new twist on the 80/20 rule. While high volume keyword conversion rates dropped significantly as positions fell, low volume keywords actually had higher conversion rates with lower positions.
Yup, you heard that one right. For low volume keywords, applying a ranking strategy diametrically opposed high volume keyword ranking strategy actually will generate higher conversions.
What this means: either the people who use low volume words are much smarter than the rest of us (in keeping product advertising costs low so that savings can be passed on to consumers). Or, they are such dim bulbs that they can’t grasp the concept of ranking order 1 to 10, 10 being the lowest. In any case, it’s worth taking a look at lower rankings for low volume keywords.
Go forth and draw your own conclusions
The data represented in these reports serve merely as guidelines for a thought progression in managing the paid search marketing process. Is it important to be number one? Of course it is, but a $10.00 click cost will not magically bring your return on investment requirement down. If $3.00 is all your brand can effectively afford, then don’t spend any more than that on a click.
Do your own tests, particularly for brand terms. My own performance comparisons suggest that brand searchers prefer official brand sites; conversion rates, along with traffic, for these terms don’t change from position one to three. Position three, four, or five, might just be better for you than position one.
Who knows you might just disprove a positioning theory. I’ve seen stranger things happen in this business and in Prague -- I'm pretty sure it was Prague -- for that matter.
You can download the Atlas information here, part one and part two.
About the Author: iMedia Search Editor Kevin Ryan’s current and former client roster reads like a “who’s who” in big brands; Rolex Watch, USA, State Farm Insurance, Farmers Insurance, Minolta Corporation, Samsung Electronics America, Toyota Motor Sales, USA, Panasonic Services, and the Hilton Hotels brands, to name a few. Ryan believes in sound guidance, creative thought, accountable actions and collaborative execution as applied to search, or any form of marketing. His principled approach and staunch commitment to the industry have made him one of the most sought after personalities in online marketing. Ryan volunteers his time with the Interactive Advertising Bureau, Search Engine Marketing Professional Organization, and several regional non-profit organizations. In his off-time, Ryan enjoys serving as Vice President at Wahlstrom Interactive.
Meet Ryan at the Kelsey’s ILM:4 Conference, November 3-5, 2004 and Ad:Tech, NY November 8-10.