Objective & Background
In December 2007 and January 2008, iProspect partnered with JupiterResearch to develop and field a survey containing questions about the behavior of search engine users. The objective of the survey was to uncover data that would enable search marketers to better understand how search engine users behave when they conduct different types of searches and are presented with different types of search results within the three largest search engines (in terms of market share): Google, Yahoo!, and MSN.
The survey focused largely on search engine user behavior as it pertains to “blended” search results (coined “Universal Search” by Google in May, 2007) that are now presented to search engine users by all three major search engines (initiated within MSN in July, 2007 and within Yahoo! in October, 2007). The results of this survey and subsequent analysis by iProspect comprise this document — the iProspect Blended Search Results Study — in which iProspect also provides actionable advice to search marketers on how to capitalize on these findings.
Beyond its focus on blended search results, the survey also sought to uncover several other search engine user behavior trends. In order to accomplish this, it included a number of questions that were also asked in previous iProspect surveys whose results were published in four prior studies: Marketing Tactics of Big Brands Not Meeting Web User Expectations (July, 2002), Searcher Behavior Shows Top Listings Are Most Important (November, 2002), the iProspect Search Engine Marketing User Attitudes Study (April, 2004) and the iProspect Search Engine User Behavior Study (April 2006). Trending information within the iProspect Blended Search Results Study is derived from these previous studies.
The panels used for each of the studies whose results are compared to the survey results that comprise this 2008 study were all different. iProspect believes that an apples-to-apples comparison of the results can still be made because in each case, the third-party provider of the panel attested to the fact that the panel was representative of the U.S. online adult population at the time the survey was conducted. It should also be noted that in each case an independent third party compiled and analyzed the results of the surveys prior to providing the results to iProspect. The methodology for the iProspect Blended Search Results Study is described below, and the methodologies used for the 2006, 2004, and 2002 studies can be found in the Appendix at the end of this study.
Search Engine Marketing Studies
Study Methodology
In December 2007 and January 2008, JupiterResearch designed and fielded a survey to online consumers selected randomly from the NPD U.S. online consumer panel. A total of 2,404 individuals responded to the survey. Respondents were asked questions about their behaviors, attitudes, and preferences as they relate to portable and home consumer electronics devices, home networks, search engine use, and data and voice services. Respondents received an email invitation to participate in the survey with an attached URL linked to a Web-based survey form. The samples were carefully balanced by a series of demographic characteristics to ensure that they were representative of the U.S. adult online population.
Demographic weighting variables included age, gender, household income, household education, household type, region, market size, race and Hispanic ethnicity. Additionally, JupiterResearch weighed the data by AOL usage, online tenure, and connection speed (broadband versus dial-up), three key determinants of online behavior. Balancing quotas are derived from JupiterResearch's Internet population model, which relies on U.S. Census Bureau data and a rich foundation of primary consumer survey research to determine the size, demographics, and ethnographics of the U.S. online population. The survey data is fully applicable to the U.S. adult online population within a confidence interval of plus or minus 3%.
In this survey effort, JupiterResearch worked with its research partner, NPD, on the technical tasks of survey fielding, sample building, balancing, and data processing. NPD is one of the largest market research companies in the U.S. and maintains a general research panel of 4 million individuals, of which 750,000 are kept "active.” The active panel receives surveys while inactive panelists are rested. This rotation in and out of active status helps keep panelists fresh and prevents burnout. Panel-based market research enables researchers to have baseline knowledge of each survey respondent and increase survey participation rates.
Search Engine Marketing Studies
Executive Summary
The findings from the iProspect Blended Search Results Study demonstrate the need for search marketers to optimize their digital image, video, and news assets to maximize the chances of these assets being displayed when search engines return “blended” search results. In addition, it also underscores the continued importance for marketers to ensure that their digital assets are found within the first three pages, if not the first page, of search results.
Prior to posing questions to respondents, the survey defined the term “narrowing options” to describe the functionality that search marketers have traditionally called “invisible tabs” or “vertical search.” In layman’s terms, it refers to the functionality that allows search engine users to limit the results of their search (prior to conducting a search) to a specialized category, such as news, images, or videos. For purposes of defining the terminology to be used through this study, findings drawn from questions that referenced the use of “narrowing options” will be expressed as “vertical search” or “vertical search results.” Findings drawn from questions that referenced “general searches” will be expressed as “blended search” or “blended search results.”
Findings & Implications
* Vertical search is not prevalently utilized by users of the “big-three” search engines (Google, Yahoo! and MSN). The survey’s results indicate that a full 35% of search engine users do not use vertical search, and 25% do not recall if they have clicked a result after having used vertical search (60% combined). A result presented after an "image search" is clicked by 26% of users (the most frequently clicked “vertical search” category). The second most commonly clicked vertical search category is "news search" at 17%, followed by "video search" at just 10%.
* In the short time that the three major search engines have provided users with blended search results (Google since May 2007, MSN since July 2007, and Yahoo! since October 2007), a greater percentage of search engine users click these "specialized" search result types within the general search results than when they perform a vertical search. Interestingly, "news" results are the most clicked specialized results within the blended search results. However, "image" results are the most clicked results when vertical search is utilized. So despite search engine users’ relative lack of use of vertical search, it is becoming increasingly important for search engine marketers to optimize their news, image, and video assets in order to be found in the blended search results.
Deeper analysis indicates that search engine users click "news" results more than twice as much (36%) within blended search results as they do when they use the vertical "news search" (17%), and click more "image" results (31%) within blended search than when utilizing vertical "image search" (26%). The extent to which they report clicking "video" results within blended search results (17%) also far exceeds the percentage of users who click a vertical "video search" result (10%). Only 19% of search engines users report that they have not clicked a news, image, or video search result within the blended search results. This compares to 35% who have not used vertical search.
* Regarding findings for which iProspect has trending information from previous studies, key among them is that 68% of search engine users click a search result within the first page of results, and a full 92% of search engine users click a result within the first three pages of search results. The importance of appearing high in the search results has steadily increased over time. We see a clear trend between 2002, 2004, 2006 and 2008 data as it relates to this finding. The data indicates that more search engine users click the first page in 2008 (68%) as compared to than in 2006 (62%), 2004 (60%) and 2002 (48%). Inversely, fewer search engine users are willing to click results past the third page in 2008 (8%) as compared to 2006 (10%), 2004 (13%) and 2002 (19%). So more than ever, it is vital for search marketers to ensure that their digital assets appear within the first three pages of search results, and especially on page one.
* In a second related finding, 49% of search engine users who continue their search process when not initially finding what they seek, change their search term and/or search engine after reviewing just the first page of search results. A full 91% do so if they do not find what they seek in the first three pages. When examining the data from 2002, 2004, 2006 and 2008, we see that more users are abandoning their query after reviewing the first page in 2008 (49%) as compared to 2006 (40%), 2004 (42%) and 2002 (28%). Inversely, fewer users are willing to continue their review of results past the third page in 2008 (9%) as compared to 2006 (12%), 2004 (17%) and 2002 (22%). Again, the need to obtain top placements on search results pages has progressively increased in importance for search engine marketers.
* Finally, it continues to be apparent that brand equity is conveyed upon companies whose digital assets appear among the top search results by roughly a third of the search engine users. In 2008, 39% of search engine users believe that the companies whose websites are returned among the top search results are the leaders in their field. This figure has grown from 36% in 2006, and 33% in 2002. Another 42% feel neutral on this question, with only 19% believing that top search engine rankings do not automatically denote an industry leader. This finding represents a significant opportunity for brand marketers in particular. Specifically, it is a convincing argument for why they should become knowledgeable about, involved in, and integrate their efforts with, search marketing. Initiatives that produce top search engine rankings can clearly help them achieve their branding goals.
Search Engine Marketing Studies
Survey Questions, Results & Analysis
Survey respondents were provided with this statement prior to being asked questions that focused on their use of vertical and blended search:
“Some search engines allow you to use narrowing options to limit your search to specific types of results. These options can be found above the main search box. For example, you can select ‘Images’ to search only within image results.”
After being presented with that statement, search engine users were asked:
1. “Thinking about a search result you clicked on in the past 6 months on Google, Yahoo!, or MSN, which of the following narrowing options did you use to reach that result? (Select all that apply)”
The results are as follows:
Click here to enlarge graph.
Analysis
From the findings of the survey, it is understandable why vertical search earned the moniker “invisible tabs.” These hidden-in-plain-sight links for images, news, and other vertical-specific digital assets have appeared above the main search box of Google, Yahoo! and MSN for a number of years. Yet over a third of search engine users do not click on them, and another quarter do not recall if they have clicked a result using this functionality.
It should be noted that as of the date of this study’s publication, Google – the largest search engine in terms of market share of searches performed – does not offer a vertical-specific search for videos on its main search page. Instead, video search is an additional click away, appearing within a sub-menu of vertical search options that is presented by clicking the “Other” link on Google’s main search page. This is one of several probable reasons why video finishes a distant third behind images and news in terms of vertical search usage.
Another reason videos are clicked infrequently compared to news and images is that video is still the newest of these three specialized search result types, so the category may have some catching up to do in terms of user adoption.
Finally, with the advent, popularity, and adoption rate of YouTube as a video-specific search engine – and its significant market share of video searches – a high percentage of online users undoubtedly choose YouTube as their source for video search. Contrast YouTube as it applies to videos with the world of image and news search – where no such dominant players exist – and it would follow that Google and the other major search engines would be utilized more frequently to search for these news and images assets than for videos.
An ancillary finding is that broadband search engine users utilize vertical search more than dial-up search engine users – with 30% of broadband search engine users conducting image searches compared to 20% of dial-up search engine users, and 13% of broadband search engine users performing video searches compared to just 5% of dial-up users. This is understandable given the inferior performance of dial-up in performing screen re-draws and in displaying image and video assets on a vertical search result screen, causing dial-up users to seek these assets less frequently. This may also speak to the level of search engine knowledge and sophistication of broadband search engine users, compared to dial-up search engine users.
Advice for Search Marketers
If the findings of this study were limited to vertical search, the takeaway for search engine marketers might be no more than to rigorously optimize images that appear on your website in order to drive visitors to your site so they can be exposed to the content of your site. And though this is still sound advice, the findings that follow pertain to blended search and prompt further actionable advice regarding the optimization of news and video assets.
Search engine users were then asked:
2. “Within the last 6 months, when performing a general search within Google, Yahoo!, or MSN (not using the narrowing options described in the previous question), which of the following types of results have you clicked on? (Select all that apply)”
The results are as follows:
Click Here for a larger version of Graph.
Analysis
When it comes to blended search results, the findings tell a different story than those of vertical search results discussed in Question #1. Roughly a third of search engine users click a news (36%) or image (31%) result after conducting a general (non-vertical) search. This compares to just 17% and 26%, respectively, for these assets being clicked by search engine users within vertical search results. Within blended search results, only 19% of search engine users report that they have not clicked a news, image, or video result (compared to 35% who have not clicked these types of results via vertical search). Also notable is the fact that 17% of search engine users click video results appearing within blended search results, compared to just 10% of users who click on a video after conducting a vertical search.
It is not surprising that news, image and video assets are being clicked by far more search engine users within blended search results than within vertical-specific searches. The most obvious reason for this is that these digital assets are presented to searchers upon their first query, enabling them to click one of these types of results without having to launch a subsequent search if they identify one of the initial results to be relevant to their search intent. In effect, the search engines are not requiring their users to overtly change their behavior – as they did when they first presented their “invisible tabs.” Instead searchers just launch a general search as usual, and these specialized results are presented to them.
Of additional interest is the fact that within blended search results, the most clicked type of vertical result is news, as opposed to images – which are clicked most frequently within vertical-specific search results. A possible reason for this would be that news-related content is fresh and current and often tied to some pre-awareness on the part of the searcher. Hence, it’s more engaging and of interest to the searcher. News headlines are also typically written to capture the reader’s (searcher’s) attention and spark interest.
Ancillary findings include: image results are clicked within blended search results by 43% of search engine users age 18-34 compared to just 24% of users age 35+; video results are clicked within blended results by 22% of male search engine users compared to just 11% of their female counterparts; and video results are clicked within blended search results by 22% of search engine users who access the Internet via broadband compared to just 9% who access it via dial-up.
A large part of the engines’ motivation for providing blended search results – aside from playing “follow the leader” – has been to try to provide a variety of result types to search engine users whose search intent is less than clear due to the ambiguity of the keyword phrase in their query. As the example below illustrates, a search for “Mount Everest” could be motivated by the search engine users’ desire to learn the mountain’s height, view a picture of it, read the latest news about expeditions climbing it, view a video taken by a climber, find an authoritative book about it, or any number of other reasons.
Click here to enlarge graph.
Because of the vague context of certain keyword phrases that are being searched upon – call them “exploration” or “discovery” searches – the three major engines tend to look at such a search as a user’s “starting point.” They fully anticipate that subsequent searches will probably have to take place for the user to find the result that is most relevant to their true intent. So the engines are now experimenting with returning various combinations of vertical digital assets, placed in a variety of locations on the search results page, and are amassing significant data on user click-through behavior.
With this user behavior data, over time the search engines hope be able to identify which types of digital assets, located at which locations on the search results page, are found to be most relevant to search engine users on a keyword by keyword basis. Armed with this intelligence, they will be able to refine the algorithms they use to determine which digital assets to present on the blended search result page to improve the relevancy of the search results for the user.
In contrast to the example above, the intent of a user’s search for “Mount Everest height” is much clearer: to “find” or “locate” a very specific fact or piece of information. Such a search typically returns only traditional Web page results and no images, videos, or news results.
Advice for Search Marketers
It should be clear to marketers that users click different types of vertical results (images, news, and video) within blended search results in far greater numbers than they do when conducting vertical searches. So where it may not have seemed worthwhile to optimize all your digital asset types to be found by vertical-specific searches – due to low user adoption for those categories – it is now paramount for marketers to do so for purposes of being found within the blended search results.
Moreover, the engines are still in the early stages of experimentation. At the time the survey was conducted, Yahoo!, for example, had only been returning blended search results for three months. So the variety of digital asset types and locations on the results page are still being tested—as are the number of results appearing on the search results page. Search engine user behavior is still being monitored. Data is still being gathered and analyzed, and algorithms are still being adjusted to focus on a variety of different “signals” associated with each result type to try to improve relevance. Marketers don’t yet know – and perhaps never will – why and how the engines include or exclude certain digital assets from blended search results, and why and how they place certain asset types in specific locations on the results page. But the bottom line is that these asset types are appearing, and will continue to appear, as a result of certain types of searches.
It’s imperative that marketers optimize all their digital assets in order to maximize their chances of garnering real estate on the blended search results page. For assets such as image and video – where once these are indexed by the search engines they have a good chance of being syndicated onto image-specific or video-specific websites (such as YouTube) – there may be legal or trademark issues to address. But even if the vague intent of “exploration” or “discovery” searches necessitates subsequent searches, the appearance of assets on the initial blended search results page can provide a positive brand impression to the searcher and help influence the clearer intent and context of the subsequent search – where assets may appear even more prominently than in the initial search results.
Search engine users were also asked:
3. “When you perform a search on a search engine and are looking over the results, approximately how many results do you typically review before clicking one? (Select One)”
The results are as follows:
Click here to enlarge graph.
Analysis
Since 2002 iProspect has been asking search engine users this question (see trending data within the chart below). The most recent findings demonstrate a continued increase in search engine users’ propensity to review only the first few search results prior to clicking one (27% of users in 2008, compared to 23% in 2006 and just 16% in 2002) as well as to review only the first page of results prior to clicking one (68% combined in 2008, compared to 62% in 2006 and just 48% in 2002).
The first three pages of search results now appear to be the “last frontier” past which very few search engine users journey – with a full 92% of search engine users typically clicking a result within the first 3 pages in 2008. In 2004, this figure was 88% and in 2002 just 81%. Viewed another way, in 2008 nearly as many search engine users review only the first two pages of search results prior to clicking one (85% combined) as reviewed the first three pages just 4 years ago (88%).
Possible reasons for this include: refinement of the search engines’ algorithms to provide and rank more relevant search results, improved searcher query refinement, increased user expectations and confidence in the ability of search engines to provide relevant results on the first page, decreased patience on the part of end users to look through larger numbers of results, the increased ability of search marketers to optimize website content to appeal to search engine algorithms, and the increased ability of paid search marketers to accurately identify user intent through the keywords they search and to craft compelling ad creative that motivates users to click.
Advice for Search Marketers
The implications of this finding should be obvious for search marketers. Now, more than ever, it is vital for a website to be found within the top few search results, or the first page of search results, or at least within the first three pages of search results. Whether this is accomplished through optimization of Web pages, or through paid search ads, the need is unmistakable. And as the findings from Question #2 indicate, with the advent of blended search results, opportunities to appear within the top search results are not limited to pages within a website. There are also opportunities for other digital assets, such as press releases, that can appear within the “news” results, as well as images, and videos, to appear within the top search results. And all of these assets can be optimized to improve their chances of being returned within blended search results.
Global Search Marketing Note
Search marketers running campaigns outside of the U.S. should take note that because the search engine optimization and paid search marketing industries are less mature in those regions than in the U.S., the user behavior trends demonstrated within this U.S.-specific finding, and the two additional findings that follow, will likely be mirrored in markets outside the U.S. in coming years.
Another factor that will contribute to this trend in markets outside the U.S. is the level of search engine user sophistication within each market – driven by the percentage of households that own computers, Internet user adoption rates, broadband availability, and users’ online tenure.
Search marketers in less mature markets should view the trends within this study as a crystal ball into the future of their markets, and take immediate steps to dominate their competition within the search results. Well optimized sites with lots of rich content and lots of links pointing to them take time to develop. Those organizations that get there first in their specific geographic markets will gain a significant competitive advantage that could potentially be maintained for years.
Search engine users were also asked:
4. “When you perform a search on a search engine and don’t find what you are looking for, at what point do you typically either revise your search, or move on to another search engine? (Select one)”
The results are as follows:
Click here to enlarge graph08_7.jpg
Analysis
This is another question iProspect has included in surveys since 2002, and its findings once again demonstrate a clear trend (see chart below). In 2008 a combined 49% of search engine users re-launch their search if they do not find a result to their liking within the first page of search results. This compares to a combined 41% in 2006 and just 28% combined in 2002. And in 2008 only 9% of search engine users wait until they have reviewed more than the first three pages of search results before they re-launch their search, while in 2002 this figure was 22%
Click here to enlarge chart.
This finding appears to crystallize three of the potential factors posed as reasons behind the finding of Question #3, namely: that over time user expectations of the search engines have increased, user search refinement has increased, and user patience to review search results for the one that meets their needs has decreased.
Advice for Search Marketers
The advice for search marketers from this finding is much the same as was provided for Question #3 – that appearing as high as possible within the search results and on as many keywords as possible – be it with pages from your website, or other digital assets such as news, images or videos – is vital to successful search marketing, and that optimization of those assets and the use of paid search ads are keys to appearing as high as possible on the search results page.
One factor that is especially applicable to this finding is that of “relevancy.” Users abandon their review of search results because their eyes do not catch content on the page that they deem relevant to their searches. So not only is it incumbent upon search marketers to optimize their sites well, and to bid appropriately to appear as high as possible within the search results, but it is also of significant importance that the title tags, page descriptions, and general page content be as focused and relevant as possible to its targeted keyword phrase. This will help ensure that the content presented to the searcher in the organic search results is as relevant as possible to the searcher’s intent. In the case of paid search ads – where search marketers have complete control of the content that appears on the search result page – well-crafted, relevant creative will help accomplish the same objective.
Search engine users were asked:
5. “Please state how much you agree/disagree with the following statement: ‘Seeing a company listed among the top results on a search engine makes me think that the company is a leader within its field.’ (Select one)”
The results are as follows:
Click here to enlarge graph.
Analysis
The final question that has been asked since 2002 is one that speaks to “brand equity.” The findings of this question clearly demonstrate the extent to which search engine users perceive that the brands whose Web pages, or other digital assets, appear among the top results on a search results page are the leading brands in their respective marketplace or category. The upward trend reflected in the chart below is clear – with an increase from 33% to 39% between 2002 and 2008 – with over a third of search engine users believing this to be true. And though there is a significant percentage (42%) that report being neutral on this belief in 2008, only 19% overtly disagree. Perhaps this 19% represents a skeptical minority who believe that anyone can “game” their way to the top of the search results without having relevant content on their site, as well as significant endorsement from other relevant websites in the form of links to their site.
The “chicken-or-the-egg” question that this finding raises, however, is whether search engine users assume that brands that are unknown to them that they find among the top search results are marketplace leaders – or whether, based on their experience as searchers of the Internet, whenever they have searched on a category they discover brands they are already familiar with and already perceive to be marketplace leaders at the top of the results.
Advice for Search Marketers
Regardless of the answer to that question, the perception that top rankings are indicative of marketplace leadership exists on the part of a third of search engine users, and less than a quarter of users disagree with the contention. So search marketers – as well as traditional and brand marketers – would be well advised to implement and integrate both paid and organic search marketing efforts as part of their overall strategy to build brand equity and increase brand awareness.
Search Engine Marketing Studies
About iProspect’s Research
Founded in 1996, iProspect is the Original Search Engine Marketing Firm. We help organizations with large, complex websites increase their online ROI and market reach through natural (organic) search engine optimization, pay per click advertising management, paid inclusion management, shopping feed management, and other related services.
iProspect has a long legacy of research and thought leadership in the search marketing industry:
iProspect’s Offline Channel Influence on Online Search Behavior Study in August 2007.
iProspect’s Search Marketer Measurement & Performance Study in June 2007.
iProspect’s Search Marketer Social Networking Study in May 2007.
iProspect's Social Networking User Behavior Study in April 2007.
iProspect’s Search Engine User Behavior Study in April 2006.
iProspect’s Natural SEO Outsourcing Study in August 2005.
iProspect’s SEO Metrics & ROI Study in August 2005.
iProspect’s Search Engine Marketer Performance Study in August 2005.
iProspect’s Natural SEO Keyword Length Study in November 2004.
iProspect’s Search Engine User Attitudes Studyin April 2004.
Searcher Behavior Shows Top Listings are Most Important in November 2002.
Marketing Tactics of Big Brands Not Meeting Web User Expectations in July 2002.
How Visible is the Fortune 100 to Web Searchers in February 2001.
Findings from iProspect research are regularly used to enhance our service offerings and to educate clients on search engine marketing best practices and industry trends. iProspect studies are frequently quoted by speakers at search marketing industry events, and by both business and trade press.
Proper attribution requires that the study is clearly identified as the “iProspect Blended Search Results Study.”
With U.S. offices in Watertown, Massachusetts and San Francisco, California, as well as offices across the globe, iProspect can be contacted at 1-800-522-1152, or by visiting www.iprospect.com.
Questions regarding this release should be directed to iProspect Media Relations Manager, Colleen Reed, at 1-800-522-1152 x1203 or colleen.reed@iprospect.com.
Click here to return to iProspect's search engine marketing library.
Search Engine Marketing Studies
Appendix
2006 Study Methodology
In January 2006, Jupiter Research designed and fielded a survey to online consumers selected randomly from the Ipsos U.S. online consumer panel that resulted in the iProspect Search Engine User Behavior Study. A total of 2,369 individuals responded to the survey. Respondents were asked approximately 25 closed-ended questions about their behaviors and preferences regarding online search, online holiday shopping, wireless services, online dating, and pharmaceutical websites. Respondents received an e-mail invitation to participate in the survey with an embedded URL linked to the Web-based survey form. The samples were carefully balanced by a series of demographic and behavioral characteristics to ensure that they were representative of the online population. Demographic weighting variables included age, gender, household income, household education, household type, region, and market size. Additionally, JupiterResearch took the unconventional step of weighting the data by AOL usage, online tenure, and connection speed (broadband versus dial-up) – three key determinants of online behavior.
Balancing quotas are derived from Jupiter Research's Internet Population Model which relies on U.S. Census Bureau data and a rich foundation of primary consumer survey research to determine the size, demographics and ethnographics of the U.S. online population. The survey data is fully applicable to the U.S. online population within a confidence interval of plus or minus 3%.
In this survey effort, JupiterResearch worked with its research partner, Ipsos-Insight, on the technical tasks of survey fielding, sample building, balancing, and data processing. Ipsos-Insight is one of the largest market research companies in the U.S. and maintains a general research panel of 400,000 households. Ipsos-Insight also has access to the Ipsos U.S. Online Panel that is comprised of two million Internet users, and provides JupiterResearch with an easy way to target and survey current online users. Panel-based market research enables researchers to have baseline knowledge of each survey respondent, to increase survey participation rates, and to permit careful rationing of survey fielding to reduce survey burnout.
2004 Study Methodology
Survey participants were recruited from a panel of nationwide participants from Survey Sampling International’s (SSI) SurveySpot, a multi-sourced panel of Internet users in the U.S. who were interested in participating in online research. SurveySpot members came from various sources, including: banner ads, other online recruitment methods, and RDD telephone recruitment. All members were recruited using permission-based techniques. SSI does not use unsolicited email in building the SurveySpot panel.
SurveySpot panel demographics were not based on predictive techniques. They were created from self-reported, respondent-specific information. This had the advantage of giving researchers greater assurance of reaching the exact targets they sought. Panel usage was monitored to prevent over-surveying as well as under-surveying in an effort to maintain panelists’ interest in participating. Panelists were also offered rewards with each survey invitation, increasing their likelihood of participation.
For the iProspect Search Engine User Attitudes Survey, a random sampling of 13,555 SSI SurveySpot members was solicited. At the suggestion of the market research professionals at Strategem Research, an independent market research firm, an incentive of $100 to 10 randomly selected respondents was offered, resulting in 1,649 responses, or 12% of those solicited. SSI invited SurveySpot members to take the survey at a website built using WebSurveyor’s online surveying capabilities.
Initial survey questions were developed by iProspect as a follow-up to, and expansion of, its 2002 survey, with the questions being vetted by the market research professionals at Strategem Research. Data was collected by the WebSurveyor online tool and results were analyzed by both iProspect’s research department and Strategem Research.
A sample size of 1,649 for a population of 170 million (U.S. Internet users as of 2004) represented a 2.41 margin of error with a 95% confidence level of a 3.18% margin of error with a 99% confidence level.
2002 Study Methodology
The iProspect studies entitled Marketing Tactics of Big Brands Not Meeting Web User Expectations (July, 2002), Searcher Behavior Shows Top Listings Are Most Important (November, 2002) are based on the results of an email-based survey hosted by Vote.com, the polling service of former presidential advisor Dick Morris. The 15-day survey ran from April 30 to May 14, 2002. The survey drew 1,403 participants from a panel provided by Vote.com of over 2 million registered Internet users. Not all questions were answered by all participants.
