Categories World Wide Web

Continued from Metadata: History and Syntax

Search engine ratings help bring in customers and revenues. To get the best ratings, use meta tags--it's that simple. Meta tag techniques that'll stack the odds in your favor (fairly, of course) include keyword lists set with NAME="KEYWORDS", descriptions set with NAME= "DESCRIPTION", and indexing controls set with NAME="ROBOTS".

When it comes to finding your Web page in a search engine, a <META> tag keyword list gives your page a definite advantage. To create a keyword list, use NAME=KEYWORDS, such as

<META NAME="KEYWORDS" CONTENT="sports books, 
history books, geography books">

Remember meta tag keyword lists are normally used in addition to those keywords the search engine gathers from the main page text. So rather than repeating keywords from the main text, concentrate on related topics or variations of the primary keywords. For example, if the keyword is computer, use variations such as PC, desktop, and notebook.

You can also use combinations of keywords or short phrases. Don't repeat a keyword several times. Instead create word combinations, such as net PC, notebook computer, and handheld PC. Watch out, though: Some search engines penalize you for repetition, disregarding the list entirely if you repeat keywords too many times. Excessive repetition of keywords is called spamdexing (as in spam indexing), and it's a bad thing. Limit the keyword list to fewer than 1,000 characters and use individual keywords no more than seven times.

Add a description to your page using the NAME=DESCRIPTION attribute. The advantage to describing pages in meta tags is that you can provide the exact description rather than having the search engine extrapolate the description from the main text. A good description summarizes the main points of the page in 200 characters or fewer. Because some search engines use page descriptions smaller than 200 characters, try to put the most relevant information first, as you see in the following example:

<META NAME="DESCRIPTION" CONTENT="PC Magazine Online is the #1 destination for technology information. We're your independent guide to personal computing and the Internet. Drop by today. See how PC Magazine can make a difference in your life.">

Because of the great increase in multilingual search engines, you may want to create language-specific versions of keywords, descriptions, and other metadata. To do this, use the LANG attribute to specify a language preference:

LANG="en"    British English
LANG="en-us" U.S. English
LANG="ko"    Korean
LANG="ru"    Russian
LANG="zh"    Chinese

You will end up with multiple (but language-specific) entries for certain meta values, such as

<META NAME="KEYWORDS" LANG="en-us" 
CONTENT="vacation, vacation books">

<META NAME="KEYWORDS" LANG="en" 
CONTENT="holiday, holiday books">

Most search engines use spiders, crawlers, or other Web beasties to index your pages. These specialized indexers find resources by crawling through the links in pages, starting at the top-level links and moving down through successive levels of links. Unfortunately, search engines may limit the number of pages they index at each Web site or they may index pages you don't want indexed, such as a page that's part of a frame set. Either way you have problems. You may have important pages that aren't indexed and that users won't be able to find easily. Conversely, you may have pages improperly indexed, so users could start their visit in places they shouldn't.

To resolve these problems, use NAME=ROBOTS, and then use the following content values to specify how the page should be indexed:

ALL Index the page and all of its links (default)

INDEX Index the page

NOINDEX Don't index the page

FOLLOW Follow the links

NOFOLLOW Don't follow the links

You can specify that the indexer should neither index this document nor analyze it for links as follows:

&l;tMETA NAME="ROBOTS"

CONTENT="NOINDEX, NOFOLLOW">

Note that the case used in this example is mandatory. According to the HTML 4.0 specification, index-related keywords and the ROBOTS name value should be specified in uppercase. An alternative to using NAME=ROBOTS in each Web page is to create a robots.txt file and place it in your Web site's main directory. The robots.txt file works much like NAME=ROBOTS but can be used to specify sets of files and directories to include or exclude from indexing.

Next: Manage Content

Published as Internet Builder in the 7/1/99 issue of PC Magazine.