A new study suggests that the overlap between search engine results is less than it was even a few months ago, and that the voices of each engine are growing even more unique.
It's easy to think of search engines as black boxes that all deliver more or less the same results. But that's not an accurate view, according to a study released today by Dogpile and researchers from the University of Pittsburgh and Penn State University.
The study looked at search results from more than 12,500 random queries on Ask Jeeves, Google, MSN search and Yahoo, and found that the overlap in first page results for these four engines was a scant 1.1% on average for a given query, suggesting that each of the four major search engines has a unique voice that's not duplicated by the other services.
The overlap between engines increases somewhat when you compare just two or three engines, but it's still less than the overlap observed in a previous study released last May.
Looking at the organic or natural listings for more than 485,000 first page search results, the study found that:
- 84.9% of total results are unique to one engine
- 11.4% of total results were shared by any two engines
- 2.6% of total results were shared by any three engines
- 1.1% of total results were shared by any four engines
The implications of these findings are significant for both searchers and marketers. Searchers relying on a single search engine are missing a vast swath of web content that they could easily find simply by trying their queries on other engines. And marketers targeting a single search engine in their optimization efforts are effectively abandoning a potentially huge amount of traffic from other sources.
Just how unique are the results on each engine? On average:
- 73.9% of Ask Jeeves first page results were unique to Ask Jeeves
- 71.2% of Yahoo first page results were unique to Yahoo
- 70.8% of MSN search first page results were unique to MSN search
- 66.4% of Google first page results were unique to Google
Effectively, for more than two-thirds of all queries, each search engine is likely to give you completely different results. So much for the black box theory, and the idea that all search engines are the same!
These findings were for organic or natural search listings, but the study also found surprisingly little overlap in sponsored listings between Yahoo and Google (Ask gets sponsored listings from Google and MSN from Yahoo). Looking at sponsored listings, the study found that just 4.7% of Yahoo and Google sponsored links overlap for a given query.
Furthermore, for about 15% of all queries, Yahoo did not return a sponsored listing when Google returned one or more, and vice versa. And overall, nearly one-third (29.6%) of all searches lacked a sponsored listing altogether in either Yahoo or Google.
This suggests that opportunities continue to abound for search marketers to gain exposure despite increasing competition among advertisers, particularly for less commonly used "search tail" terms.
Dogpile's motivation for doing this research is to demonstrate that searchers have a better chance of finding what they're looking for using a meta search engine Dogpile.com. There's merit to this argument, especially with the recent interface changes Dogpile made making it easy to compare the results between the four major engines (Dogpile added MSN search results today).
Beyond a marketing effort, however, the new report goes far beyond the earlier study, presenting a comprehensive summary of methodology, analysis and an excellent discussion of assumptions made and conclusions drawn. For example, the study looked only at the first page of results because the majority of search activity (89.8%) takes place on the first page, according to a log file analysis by Infospace, Dogpile's owner.
This report also makes a clear distinction between natural or organic search results and sponsored listings, something that was not included in the previous report. On balance, the report provides enough background and raw data to allow you to reach your own conclusions apart from those presented by Dogpile.
It's an excellent read, and well worth the time spent digesting its findings. Search marketers will find lots of useful nuggets for modifying or adapting campaigns based on actual searcher behavior. And even if your interest in search engines is limited to finding information, the eye-opening statistics on result overlap may encourage you to consider using multiple engines, whether via Dogpile's meta search engine or simply by visiting multiple sites.
Dogpile has made two versions of the report available for download. Both are PDF files.
Different Engines, Different Results - 30 page PDF white paper
Executive Summary - 1 page PDF
Separately, Dogpile has relaunched its "missing pieces" tool as the Dogpile search comparison tool. This interactive tool lets you enter a query and then displays unique and overlapping results on a Venn-diagram, and tells you exactly how many unique and overlapping results are found on Yahoo, Google and MSN. You can also compare Ask Jeeves results using a drop-down menu at the top of each circle. It's a fascinating way to test search engine overlap yourself.
NOTE: Article links often change. In case of a bad link, use the publication's search facility, which most have, and search for the headline.