Or, searching is really all about finding, first of three articles

How search engine marketing tools can work for you: or, searching is really all about finding, first of three articles

Terry Brainerd Chadwick

How does one go from being a librarian and independent information professional to a specialist in search engine optimization (SEO) and marketing? In my case, it was almost inevitable.


As the Internet became more and more of a repository for crucial information, an increasing amount of my work involved sifting through hundreds of Web sites, searching for pertinent information on a wide variety of projects in a very short time. Because it was so hard to find the information on many of those Web sites, my frustration kept growing. I knew there were Web sites out there with great information–I just couldn’t find them. When I did find a Web site, I couldn’t find the information I was looking for in it.

I’m a searcher. I know how to search and where to find information. I understand how search engines work. The problem? The companies and organizations behind these Web sites didn’t understand how to have their information found by the people who were looking for it. Shortly after this realization, I started working on search engine promotion and optimization. So what does a search engine optimization specialist do?

I am currently the SEO specialist at EBSCO Subsidiary Web Services (ESWS), a department within EBSCO Industries Inc. that provides Web design, development, maintenance, and Internet marketing for the company’s 20-plus subsidiary divisions. These Web sites include businesses involved in real estate, hunting and fishing supplies, office and retail products, promotional displays, and publishing. My responsibility is to develop the sites into search-engine-friendly, easily indexed, highly ranked resources. I am also responsible for making sure the Web sites easily convert visits into leads and sales.



When I start on a new project, I first conduct an SEO survey. Questions include the following:

* What are your goals for the company, its products, and the Web site?

* What are your expectations and how will you measure the success of the project?

* Who is your audience? Who are the customers and competitors?

* How do you describe your products and services (i.e., what keywords would you use to describe them)?

* What kind of Internet marketing have you already done?

* Is your Web site database driven? What kind of server does it run from? Does your site use JavaScript, Flash, PHP, CGI, ASP, .NET, or other scripting and programming languages?

Using the information gathered in the survey and continuing to ask questions as needed, I start my research. This includes reviewing the Web site and other company marketing materials to see how they describe their products and services, reviewing competitors’ Web sites, and conducting a keyword search and site-saturation analysis. I pull this information together to create an evaluation of the site’s search engine friendliness.

Search engine marketing (SEM) is more art than science. Each search engine has its own formula for ranking search results. These algorithms change often, primarily to make results more relevant to the searcher but also to derail spammers who try to use the algorithms to make their pages rank high regardless of their relevance. There are, however, a few general best practices in search engine marketing. Most of these are useful for anyone who wants to have a user-friendly Web site.

1. Use unique and descriptive titles on all pages.

2. Use plain ASCII text on your Web pages.

3. Use alt text to describe images and nontext elements.

4. Include a keyword-rich text description at the tops of pages (which the search engines read).

5. Provide unique descriptions and keyword metatags on every page.

6. Use text links with keyword-rich descriptions in your navigation and content.

7. Add relevant, high-quality links from other sites to your site.

8. Adhere to Web standards; follow usability/accessibility guidelines.

9. Use keywords in URLs.

10. Avoid using redirects or refreshes.

11. Do not duplicate pages, sites, or content. If you have similar/duplicate pages or sites, choose one for the search engines; exclude the search engines from the others.

12. Make sure your dynamic pages are search engine friendly.

13. Restrict the use of splash or doorway pages (pages that are graphics/flash only or pages filled with keywords for search engines, not humans, and that redirect to the real Web site).

14. Do not cloak (show the search engines pages/information that people won’t see).

Major SEM Tools and How I Use Them

There are many tools to help you optimize Web sites. They can be grouped into four categories: (1) keyword research and analysis; (2) Web site saturation and popularity; (3) back end tools (including Web analytic tools and HTML validators); and (4) Who Is tools.

Keyword Research and Analysis

The three major goals of an SEO project are to (1) make sure the site can be indexed in the search engines; (2) find the most relevant and popular key terms and phrases for the site and its products; and (3) use those key phrases on the site in a way that will generate and convert traffic.

Remember that the “most relevant” keywords aren’t necessarily the ones favored by the industry. You need to determine what words people use to search for your products and services. Researching how high your company’s Web site ranks on a given search engine is useful only if you are using the same keywords that potential customers will use. Numerous tools are available to help you capture actual searches, giving you relevant estimates of the popularity of various words and phrases.

Wordtracker[TM] (www.wordtracker.com) is my favorite keyword research tool. Enter any keyword or phrase to determine how many people have searched using these exact same words. Wordtracker reports the number of searches on the phrase in the past 60 days; the estimated number of searches per 24-hour period; the number of Web pages containing the phrase; and the most effective key phrases based on the popularity of the word and the amount of competition.

A limited trial version of Wordtracker is available (AltaVista is its test database). The fee-based version includes major search engines, pay-per-click engines, and directories. Pricing is available on a daily, weekly, monthly, quarterly, and yearly basis. One-year subscriptions are approximately $260.

Wordtracker provides data from the full database or by individual search engine. It currently offers keyword popularity information for the following search sites:

Google Yahoo Directory

MSN Open Directory

Yahoo Looksmart

Lycos Overture

AltaVista FindWhat

Teoma Searchboss

Hotbot Bay9

AlltheWeb GoClick

Reports from individual search engines contain information on the number of competing pages and keyword effectiveness.

One reason I like Wordtracker is that it gives you a list of 300 related terms and thesaurus listings for the keyword you test. You can use the popularity side of Wordtracker by directly transferring a term from the Related Words list on the left to the Popularity Search Box on the right. You can also enter a term into the Popularity Search Box to find out how many times it appears in the Wordtracker database. This database includes all searches conducted within the past 60 days.

Wordtracker provides a variety of search options, including these:

* Compressed search. Removes unnecessary characters (+, -, “) and compresses all forms of the term to lowercase. For example, “POKEmon” and “+ pOkEmon” become “pokemon.”

* Simple search. Finds the term with words before and after (also taking into account capitals and the plural form). For example, enter “hello” and get “when i say hello,” “hellos,” and “hello dolly,” but not “othello.”






* Comprehensive. Allows word stemming both left and right of the term. For example, “hellontheroad” on the right or “othellos” on both left and right.

* Exact. Returns all the upper- and lowercase combinations for the keyword you have selected. For example, if you enter “HELLO,” you will get back “HELLO,” “Hello,” “HelLo,” and so on.

* Precise. Similar to the Exact Search, but returns just the term selected, compressing all its upper- and lowercase forms into one figure

* Overture. Uses the Overture Keyword Selector Term Database (see below for description)

Choose the search type based on the search engines you’re targeting rather than the keywords you’re targeting. For instance, when targeting search engines that place different weights on uppercase vs. lowercase words, use Simple Search. The Comprehensive Search can be very useful in domain name research. The Compressed Search is the default search type and the one I usually use.


While Wordtracker is not the only keyword popularity tool available, it is one of the most comprehensive. Its main weakness is that it doesn’t use searches directly from the search engines it covers. Instead it uses searches from the metacrawler Web site Dogpile that searches most of the major search engines. Although this is a disadvantage in some ways, it is also a strength. Search term popularity numbers drawn directly from the major search engines include all the searches done by automated agents (also known as spiders and robots [bots]) to test search engine rankings. These bots launch searches on hundreds or thousands of keywords to check where sites rank. They’re not “real” searches done to answer questions. The bots don’t test rankings on metasearch engines, so the Dogpile results avoid the inflated effects of automated ranking tests.

The bottom line? Any tool you use will provide only estimates of what you can expect for your site and its products. If you use more than one tool, you can get a very good idea of what the most popular words are for each major search engine.

Overture, Yahoo’s paid placement service (www.content.overture.co/d/USm/ac/index.jhtml) provides several services to help its customers make the best decisions on choosing keywords and on pricing the bids they make for placing their ads in the Overture network. Two of these services, the Overture Keyword Selector Tool and the Overture View Bids Tool, are available for free. While these tools are designed for people doing paid placement campaigns, they are also useful for search engine optimization.

When you use the Overture Keyword Selector Tool (http://inventory.overture.com/d/searchinventory/suggestion), search data are drawn from actual searches from a previous month from the Yahoo family of databases. When you search a term on the Overture paid placement service, it covers both singular and plural terms of the word as well as common spelling variations.

The tricky part is that this doesn’t happen all the time. When I search for “children’s furniture,” the responses all come back with the singular “child” (e.g., the first response is “child furniture” with 107,894 searches for the month of October). I get the same response when I search for “child furniture” or “kids furniture.” Yet when I search for “tree-stand” (1,732 searches), I get different results than when I search for “tree-stands” (7,062 searches) or “tree stand” (4,410 searches) or “tree stands” (11,299 searches). Another downside of Overture’s combining of tenses and forms is that people often search for the same subject using similar terms or different forms of the same terms (as in the examples above). A very important part of SEM is determining which form of a word is used most often. Based on the numbers in this example, consider what would happen if you’d optimized your site for “child furniture.” You’d be missing out on a lot of searches!

The Overture Keyword Selector is limited in that it doesn’t show you multiple search engines or the keywords your competitors are using. Its advantage is that the data are compiled from actual searches from the Yahoo family of databases, not from a metacrawler as with the Wordtracker data. The Overture Keyword Selector is designed to help users of the Overture paid placement program. I use this tool as another way to check the overall popularity of terms.

The Overture View Bids Tool (http://uv.bidtool.overture.com/d/search/tools/bidtool) doesn’t really show keyword “popularity.” It does, however, show you who your competitors are and what keywords they think are important, which makes it an excellent competitive intelligence tool.

Google’s equivalent of the Overture Keyword Suggestion tool comes in a number of forms. The free, don’t-need-to-register version of the Google AdWords Suggestion Tool (https://adwords.google.com/select) gives you expansions on suggested terms without providing an estimated number of searches. To get more information from AdWords, you need to register and set up an ad campaign, although you don’t need to actually launch the campaign or make any payments.








Like the Overture Keyword Selector, AdWords is designed for paid placement, not search optimization. However, its Traffic Estimator feature determines the expected number of clicks per day for a particular ad position, providing a nice counterpoint to Wordtracker’s and Overture’s expected number of searches per day. Just because someone searches on a term it doesn’t mean they click through to a site. This option helps in budgeting a paid placement program. It also assists in figuring out how many people you can actually expect to visit your site for any given keyword and ranking.

Although I use this tool primarily for paid placement, I also use it the same way I do Overture for search engine optimization, as a tool to check the overall popularity of terms.

Trellian’s Priority Submit Search Term Research site (www.prioritysubmit.com/research.html) has a number of excellent optimization tools. Some of the tools are free with registration, but the unique ones are fee-based. The keyword analysis tool, while limited in its free version (you get only the first 20 results), has many valuable features. The database includes information compiled from 37 sources, including major international, pay-per-click, meta, and regional search engines. The Related option performs a lateral search to identify related terms, and the Overture option enables you to use the Overture search term database. You can also see the most popular and recent searches from the huge database. A unique tool shows seasonal search-trends, which can be invaluable if your products are seasonally based and you are searching off-season. Also unique in the paid service is an advanced search syntax that lets you exclude words from a search, include misspellings, and import keywords from metatags. I don’t have access to the paid version of this tool, but we are considering getting it because of the unique aspects of the service.

Many searchers misspell the names of products, companies, and people. If people commonly misspell your company name, they may not reach your site when they search for you. The SearchSpell Typo Search Tool (www.searchspell.com/typo) shows the top typos for any term entered.

Clicking on a result takes you to a Google search on the typo to help you determine the Web page popularity of any given typo.

If you use SearchSpell with Wordtracker, you can see how often people actually search with these misspelled words. You can use this information as a guide in choosing alternate domain names and meta keywords. You can also use the information to develop innovative ways to bring these people to your Web pages.

The Referrer segments of Web site log files are an important part of keyword analysis. They show what keywords people are using to actually get to your site. As I’ve optimized sites in the past, I’ve frequently discovered valuable keywords in the log files that aren’t listed as popular in my initial research. If you are getting a lot of Web site traffic based on a certain term, you don’t want to eliminate that term during site optimization just because a third party tool says it isn’t popular.

Almost all Web traffic or analytics programs have the capability to show search engine referrals. This is an excellent way to see which search engines send traffic to your Web site and to find out which keywords were used. If you aren’t running paid placement programs, it shows the keywords you rank high enough for people to find your Web site in organic listings (unpaid listings with rankings determined by search engine algorithms). If you are running paid placement programs, those keywords are usually intermixed with the organic in search term referrals. However, some analytics programs differentiate between organic and paid search listings. WebSideStory’s HBX[TM] is one of these.

The information in Figure 21 tells me that the most popular way to reach the Siegel Display Products Web site is by searching on a variation of the company name through organic listings and that paid placement is a major factor in bringing people to the site on product type searches.

The information in Figure 22 shows that people are finding Siegel Display through organic listings for the term “literature holders” on Google, but that paid placement ads are primarily responsible for boosting site traffic from other search engines receiving the same search term.

Wordtracker said that people search on the phrase “literature holders” about 38 times a day. Of this figure, 13 searches a day are on Google. This means that Siegel Display has a daily click-through rate of 6 percent overall and 8 percent on Google, which is considered pretty good. Although “literature holder” is not the most popular keyword for Siegel’s products (“magazine rack” is searched about 289 times a day), it is a term I want to keep optimized for the site, working to increase organic rankings on the other search engines.

A good analytics tool will also show you the terms that convert to orders and sales. With this information, I can emphasize converting terms more than nonconverting terms when optimizing the Web site, or I can increase efforts to improve conversion rates on the popular but nonconverting terms. “Literature holders” converted 1.74 percent of the time in August 2004 for Siegel Display, compared with “magazine racks,” which had a conversion rate of only 0.53 percent. This information, combined with other information tracked by the analytics programs, helps me determine which keywords and Web pages are working well and which need to be bolstered by optimizing content.


Figure 3

Wordtracker Search Options: Top 10 Results for “start”

Compressed Count Comprehensive Count

we didn’t start the fire 641 lets get it started 1084

head start 573 let’s get it started 1002

start using cgiproxy 472 remote car starter 882

start internet business 386 starting over 792

start an internet business 383 the starting line 785

internet business start up 355 remote car starters 739

how to start a business 354 starting a business 626

start your own business 340 Startrek 606

start your own internet 338 we didn’t start the fire 420


remote start 320 starting a small business 402

Exact Precise

Keyword Count Keyword Count

start 195 start 195


Start 29

Overture (actual queries

for past month from

Overture Database, not

Wordtracker Database) Simple

Keyword Count Keyword Count

starting a business 56579 we didn’t start the fire 420

start 43052 start using cgiproxy 387

dsl start verizon.net 23964 start internet business 386

welcome start 20320 start an internet business 383

head start 18031 internet business start up 355

start new search 17590 Head start 353

start your own business 12352 Start your own internet 338


how to start an internet 11840 how to start a business 314


remote start 10645 how to start an internet 311


start using cgi proxy 10551 remote start 297

Figure 10

Wordtracker Exact Search on Expansions on Variations of Child Furniture


Database Predicted

Keyword Count Searches/Day

Childrens furniture 1060 965

Kids furniture 1055 960

Children furniture 322 293

Kids Furniture 76 69



Childrens Furniture 28 25

Child furniture 23 21

childs furniture 22 20

Kids furniture 15 14

Childrens furniture 12 11

Child Furniture 9 8

Children Furniture 8 7


Child’s Furniture 0 0

Children’s furniture 0 0

kid’s furniture 0 0

Figure 19

Number of Pages on Google with Misspellings of Siegel

Results 1-100 of about 161,000 for sigel

Results 1-100 of about 80,200 for siegle

Results 1-100 of about 128,000 for siegl

Results 1-100 of about 1,230,000 for segel

Results 1-100 of about 143,000 for seigel

Figure 20

Wordtracker Exact Results for Misspelled Words


Database Predicted

Keyword Count Searches/Day

Seigel 24 22


Seigel 6 5

Sigel 5 5

Siegle 5 5

Segel 0 0

Siegl 0 0

Figure 21

Excerpted Example of HBX Top 10 Search Keyword Referrals

Search Keywords Visits Paid % Paid Organic % Organic

Magazine racks 187 179 95.72 8 4.28

siegel display products 182 0 0 182 100

siegel display 181 0 0 181 100

Magazine rack 172 170 98.83 2 1.17

Pegboard hooks 122 29 23.77 93 76.23

siegel displays 79 0 0 79 100

www.siegeldisplay.com 79 0 0 79 100

Literature holders 74 28 37.83 46 62.17

sign holders 73 61 83.56 12 16.44

Pegboard accessories 69 4 5.79 65 94.21

Total visits 13,088 1,218 471 38.66 747 61.34

Figure 22

Excerpted Example of HBX Search Engine Referrals for Keyword Literature


Detail: literature holders

Search Engines Visits Paid % Paid Organic % Organic

MSN 31 24 77.41 7 22.59

Google 30 0 0 30 100

Yahoo 10 3 30 7 70

AltaVista 1 1 100 0 0

Total 72 28 38.88 44 61.12

RELATED ARTICLE: How special librarians can use keyword research tools

* To find out how your competitors use industry terms, check the standard search engines. To learn how potential customers search for your products and services, use Wordtracker or another keyword popularity tool.

* Overture’s View Bids Tool is useful in gathering competitive intelligence. It tells you which keywords your competitors consider important.

Terry Brainerd Chadwick is a former corporate librarian and independent information professional turned search marketing specialist. She became involved with search engine optimization and Web design 10 years ago, after becoming frustrated with searching for information on hundreds of Web sites a day for her competitive intelligence and market research projects. She is a past president of the Oregon Chapter of SLA.


COPYRIGHT 2005 Special Libraries Association

COPYRIGHT 2005 Gale Group