Section 3. How Do Visitors Arrive at the Web Sites and Who Stays?
Users of Public Reports of Hospital Quality: Who, What, Why, and How?
Knowing more about how visitors arrive at the Web sites can guide efforts to increase Web site traffic. Two useful traffic metrics from Web analytics are: overall traffic, represented by number of unique visitors; and bounce rate, the percentage of visitors that view only one page before leaving the Web site. Lower bounce rates indicate a more engaged audience. Overall traffic can be tracked over time to assess how marketing or Web site changes affect traffic, and bounce rate can be tracked to gauge fluctuations in visitor engagement.
Sources of overall Web site traffic fall into three broad categories: search (traffic from search engines), referrals from a link on another Web site, or direct entries of the Web site URL into the Internet browser bar or clicks on a direct link in an E-mail or word processed document. Traffic can be increased from each of these sources, but the methods of achieving those increases vary by source.
There was wide variation in the proportion of traffic types for each Web site and in the numbers of visitors arriving from each source. Table 6 shows the variation among the sites in traffic from each source. Percentage of traffic arriving from searches ranged from 3 percent to 89 percent among the sites, while percentage of traffic arriving via referrals ranged from 4 percent to 84 percent, and percentage of direct traffic ranged from 6 percent to 61 percent.
Sites with a higher percentage of search traffic also had higher total traffic. In large part, this is because the maximum traffic generated through search (two sites had over 400 visitors per 100,000 Internet households from search) is much greater than the maximum traffic generated through either referrals (maximum 122 visitors per 100,000 Internet households from referrals) or direct access (maximum 80 visitors per 100,000 Internet households from direct).
Table 6. Number and percentage of visitors from search, referral, or direct* sources
|Web Site Number||1||2||3||4||5||6||7||8||9||10||11||12||13||14||15||16|
|Search visitors per 100,000 Internet households||0.03||3.4||5.6||5.3||15||14||25||28||15||36||45||22||12||175||444||424|
|Percentage of all visitors arriving via search||3||35||19||12||31||23||42||39||15||33||33||16||8||46||89||83|
|Referral visitors per 100,000 Internet households||1||2.2||14||17||19||29||24||15||22||32||18||61||91||122||27||19|
|Percentage of all visitors arriving via referral||84||23||46||41||40||48||40||22||23||29||13||43||64||32||5.4||3.8|
|Direct visitors per 100,000 Internet households||0.15||4.1||10||20||14||17||11||28||58||43||73||58||39||80||30||65|
|Percentage of all visitors arriving directly||13||42||35||47||29||29||19||39||61||39||53||41||27||21||6||13|
Note: Search is defined as traffic arriving from Web search engines (e.g., Google, Yahoo, Bing); referral is defined as traffic arriving from a link on another Web site; and direct is defined as traffic arriving either by direct entry of the Web site URL into a Web browser bar or clicks on a direct link to the Web site URL in an E-mail or word processed document. Percentages may not add to 100 due to rounding.
Only half of the Web sites are structured to allow search engines or Web analytics programs to identify hospital quality pages separately from the home page or other general pages. Web pages are programmed with "meta tags." These are not something visitors are likely to notice but are used by search engines and Web analytics programs to identify page content. However, it is possible to create a Web page in which the content can change (for example, from what looks like a page about one hospital's quality to a page of another hospital's quality) without the meta tag changing. If pages are created this way, search engines will not be able to find the content on these pages, including any hospital or performance metric specific information.
If meta tags stay the same while content changes, the Web site sponsor will have negative consequences. The search engine will not show the Web sites in the search results for any specific content that has not been tagged, and the sponsor cannot use Web analytics to determine which pages are most popular with visitors (that is, are viewed most often). In addition, higher traffic Web sites were structured to change meta tags between separate content pages, suggesting that doing so increases traffic in practice, not just in theory.
Most visitors using a search engine were searching for the Web site name itself or for a hospital. For most Web sites, the name of the Web site was the most common search by which visitors arrived. Very few visitors arrived through a search for a medical condition (this figure varied from 0% to 9% among the Web sites). In addition, few arrived through a search for "high quality hospitals" or "hospital performance."
Some site sponsors have successful partnerships that drive referral traffic. For example, the Web site that had the largest proportion of consumers (81% of survey respondents) also had a large percentage of traffic (64%) from referral sources. The two top referral sources for the site were a county government (".gov") Web site and the affiliate CVE Web site. This Web site also had a few write-in comments about a health incentives program offered by one of its partners that rewarded employees who visited the site. Not surprisingly, this site also had high overall traffic.
- Many sites excel at garnering traffic from one source (search, referral, or direct), but no site is at the top in all three sources. Therefore, all sites have opportunities to improve.
- To date, improving search traffic has been the most effective method of increasing overall traffic. Thus, search engine optimization is a critical activity for all Web site sponsors.
- It seems that there is limited competition for individual hospital names in search engines, compared to medical condition searches. Making hospital names easy for search engines to find through effective use of meta tags may be a good way to increase visits to a site.
- Our findings suggest that focused marketing can increase traffic. Since many of the top keyword searches are for the name of a specific site, it seems likely that many people who search already know the site for which they are searching and are responding to some form of marketing or advertising. This was true for sites that had high percentages of consumers and for sites that had high percentages of health care professionals, implying that marketing can increase traffic from either group. For consumers, this is likely to be general marketing (such as radio campaigns, in which the consumer learns the site name and types it into a search engine or browser bar). For health care professionals, direct traffic can be increased through list serves or other electronic means of sharing the link to the site, or through distribution of documents that contain the link.
- Conversely, our findings suggest that content about specific diseases (medical terms and conditions) does not generate much search engine traffic on these Web sites.
- This could be because people are not searching online for information about diseases very often. Much more likely, vigorous online competition for the market segment of consumers with specific diseases means that public reports of quality do poorly compared to other sites on searches for these terms.
- It remains to be seen whether public reporting Web sites could compete more effectively for these searches if they improved search engine optimization.
- Collaborative work with other organizations can increase a Web site's traffic and help gain footing in a targeted market.
- There is evidence from one Web site that incentives work to get individuals to view hospital quality information.