Search providers had the best retention rates, with an average of 71 percent of June visitors at home retuning in July among the three most popular search engines. Google search led with a retention rate of 79 percent, Yahoo Search was second at 69 percent and MSN/Windows Live Search had 65 percent. Visitor retention rates were a little higher with the work audience, who had an average of 76 percent among the three leading search providers.

Nielsen noted that a large portion of visitors went to more than one of the three search sites in July. MSN/Windows Live Search had the highest audience overlap with 84 percent of its unique visitors also going to Google, Yahoo or both. Yahoo had a 78 percent audience overlap with Google and MSN. Google had a 63 percent audience overlap with one or both of its competitors.

"These search providers are doing a good job of bringing their visitors back month after month, which is an important sign of the health of their customer relationships and a measure with significant financial implications, given the expense of acquiring new customers," said Ken Cassar, chief analyst, Nielsen//NetRatings.

The most popular three sites in the career development category were CareerBulider, Monster and Yahoo HotJobs. They had an average month- over- month retention rate of 38 percent at home and at work it was 44 percent. Monster.com had the most audience overlap with the other two, at 50 percent. Yahoo had a 45 percent overlap, while CareerBuilder was 34 percent.

The top sites in the multi-category travel, Expedia, Orbitz and Travelocity, had the lowest average month- over- month at home visitor retention rates at 31 percent. At work it was a little higher at 35 percent. Overlap was similar between the three sites at 59, 58 and 51 percent for Orbitz, Travelocity and Expedia, respectively.

"Search is a daily online activity for most Web users, so search providers can rely to a certain extent on frequency of use as a driver of repeat visitation," said Cassar.

"Career development and travel sites are used less frequently, so their challenge is to be top of mind when it comes time to look for a new career opportunity or plan a trip."

The Google CPG blog cited a study by aQuantive’s Atlas Institute about the value of overlapping ads. By running ads in multiple points of contact – publishers, portals, and ad networks – Atlas found 67 percent of people who took action did so after encountering multiple ad placements.

Google’s Emma McKeithen noted how they do this for Google’s CPG clients, and gave an example of how a given person might navigate the Internet on a typical day:

During the course of a day, a user might visit a sports site, a social network, a photo-sharing site, check Gmail (perhaps repeatedly) in addition to going elsewhere online. The goal of overlapping campaigns is to maximize these touchpoints without increasing frequency to the point where an ad placement no longer delivers returns for the marketer.

Atlas did caution that overlap’s impact on conversions will “differ greatly from advertiser to advertiser as there is no consistent correlation between reach and conversion overlap.”

By Atlas’ notes, overlap seems to work best for branding, and requires a targeted reach to be most effective. A danger exists for overexposure of the brand through such overlap, which means advertisers will want to watch for opportunities to swap overlap for superior demographic reach.

"This study reinforces what we at InfoSpace have long known – often users do not find the results they need with any single search engine. Metasearch offers the most robust and efficient search solution to meet their needs," said Rod Diefendorf, vice president of online and local search at InfoSpace.

"With less than one percent overlap in first page results, there is great value to using a metasearch engine like Dogpile.com to quickly comb through multiple search engines at once for the most relevant results."

The study found that only 0.6 percent of 776,435 first-page results were the same on all top four search engines. Between 38 and 46 percent of all searches did not produce clicks on first- page results.

On average searchers use three search engines a month. Search result rankings were different on all the major search engines with just 3.6 percent of the number one ranked non-sponsored search results were the same on all search engines in a given query.

"There is a perception that most search engines function similarly and deliver the same results, but that is not what these result show," said Dr. Jim Jansen of The Pennsylvania State University.

"In fact, the engines are increasingly diverging in their approach to searching the Web. For some searches, users are obviously missing quality results."