What are Meta Tags Anyway?
I have been building database-driven web applications for years now that have resulted in the production of thousands of web pages for sites I run or sites owned by clients. As a natural result of my work, I have had to naturally acquaint myself with the trade of Search Engine Optimization (commonly referred to as SEO). Search Engine Optimization is the art of making web pages friendly to search engines to increase traffic (and to increase an abstract concept called Page Rank).
While SEO has always played a peripheral role in what I do, it became especially important in the last several months in my personal life since a company called Cambridge Who’s Who is suing me for $7 million dollars for, among other things, getting traffic from the "inclusion” of "metatags” in my site. (See Paragraph 28 of Cambridge Who’s Who Publishing, Inc VS. Shawn Olson filed in the United States District Court Eastern District of New York on December 27, 2006, case CV-06 6770).
Of course, the whole case is a harassing sham. The fact that the lawsuit exists shows some of the ignorance about web technologies of our age even amongst professional lawyers and judges. (It also shows the extent that one company will go to use a lack of understanding to try and bully the free speech out of this online journalist.)
In their naked form, meta tags are nothing more than abstract ways to classify a document (in this case, a web page). One common meta tag is the "description”. The description is usually a single sentence or two that briefly describes a document. Instead of reading an entire document to know what it is about, you "should” be able to read its description and get an idea of what the document is about. The description is intended to give the potential reader a clue as to whether or not the current document is the one he/she is looking for.
Meta Tags are not magical. In fact, they are much less mysterious than they may seem to the average web browser and novice webmaster. In the current state of the Web, you cannot insert a phrase into your meta tags and magically appear as the number one result for those phrases in search engines based on those meta tags alone*. In fact, as it stands, meta tags (especially those that are not openly visible to web browsers) don’t seem to factor in much at all to search engine results.
Most modern search engines do not pay much heed to meta tags. Instead, search engines calculate relevance between a search phrase and a web page based on equations that factor page content and the content of in-bounding links to a page above everything else. Some meta tags are factored in, and some are ignored… but they do not equate for much of the total value. A page’s title (which is visible in browsers and search results) and its description (which often appears in search results) seem to be the only two pieces of data in the HTML head that is factored into search engines like Google. Meta tags such as keywords are not factored into results due to a history of keyword abuse in the early days of the internet and search engines.
The reason that the keywords meta tag is ignored by most search engines is that it is simple to place words that are irrelevant to a page’s actual content as a tactic to get traffic. For example, in the early days of the internet, you might insert a phrase like "transformers” into the keywords of a page about "politics” for the simple purpose of filtering traffic to your site regardless of the searcher’s intent. Search engines long ago figured this out, which is why most search engines have stopped factoring in keywords.
These facts force you to ask a key question: Why use meta tags at all if they do nothing? That question seems like an obvious next step… but it is a little off base because the above discussion did not say that meta tags (specifically keywords) do nothing. What was stated is that they are factored very lightly by mainstream search engines. But meta data can be used in different ways.
To explain the value of meta data, consider meta data as used on your personal computer. Imagine you have a folder with 10,000 photos you took with your digital camera over the last few years. If you are like many people, you may simply have dumped them into a single folder. This is fine… but what if you suddenly want to find a specific photo in your collection? If you have to preview each individually, you might take a week to find the photo you want. Meta Data can make your search easier—you can suddenly sort images by their dates if you want and find all images you took in December 2006 (effectively narrowing the search and optimizing your time). In this case, the date an image was taken is attached to a photo in the form of Meta Data—it is information about a file that is not essential to the file but helps explain the file. With digital images, there is a wide range of meta data that you can use to sort images, including their names, sizes, color depth, shutter speed, camera used, etc. Some media asset software packages even allow you to create your own fields for indexing images, such as adding keywords, topics, colors, etc.
The above example is a return to Searching. But the difference between the two kinds of searches is obvious—in one case the search is made by potential visitors on third party sites while the images where made by you on your own files. It is easy to see why someone would want to incorrectly categorize a page for third parties (to increase traffic) versus incorrectly categorize your own data for internal purposes.
Regardless for the potential for abuse, internal and third-party programs can utilize data contained in a web page’s meta data. In terms of web pages, you might insert data in keywords that allow simple crawlers to index the topics of a page without having to analyze the full content. You might make such a crawler (similar to a search engine crawler) to analyze a set of websites for which you are acquainted with but do not have full control of (such as in affiliate programs, etc) and that you trust to provide valid meta data descriptions. You might even make custom meta tags for web pages that are used by your system’s software—but that search engines will ignore!
There are many legitimate uses of meta data in web pages. The point here is that whatever you put into the meta tags will not likely give you much of an edge in search engines—and if you place irrelevant content in your meta tags, you might even be penalized by search engines for trying to manipulate the system. If a page is about golden retrievers, there are legitimate reasons to insert "dogs, canines, pets” into the keywords because those words legitimately describe aspects of the page… while phrases like "nascar” would be wrong for that page. Either way, inserting Nascar into the keywords would not give the page any results for "nascar” in search engines unless the page was about "nascar dogs” or if tons of Nascar sites linked to the page—and again, this is because of page content and in-bound links, not because of keywords.
The use of the "keywords” meta tag is of little SEO value. As such, I have always made sure that the keywords I place in my pages actually do relate to the content in a page—for the very minute SEO value they potentially have and for internal software I intend to build for sites using the Webonizer Content Management System. As such, I see the use of the keywords meta data as an internal tool for sorting data. (In the case of Webonizer, the keywords of an article are used for internal fulltext searching and as keyword links in internal search results pages.) The description meta tag probably provides the greatest external SEO value… but anyone is going to see that in a search result and quickly notice if the page is legitimate or not—and with modern search engines, it is the whole document that gives value to any search phrases, not the description*.
An interesting thing to note in regards to the Cambridge Who’s Who lawsuit against me is that their claim implies that it is unethical for me to include the phrase "Cambridge Who’s Who” in the meta data of my articles about them. If this is true, then major news providers themselves are being "unethical” by providing their subjects in their meta tags. For example, I did a search for "enron scandal” on the following major news sites and you can see the usage of keyword meta tags being employed from articles on those sites at the time I published this article (click specific links and hit View>Source in your browser to find this information).
- CNN.com : "Corporate Governance , Mark Foley , Enron Corporation , The Boeing Company , White House , Ohio , Careers”
- Fox News : "enron, fraud, skilling, business & money, business news, financial news”
- New York Times: "”
- Time: "George W. Bush, President, White House, Enron, energy company, stock value, Cheney” (article now gone)
- BBC: "BBC, News, BBC News, news online, world, uk, international, foreign, british, online, service”
Obviously, I am not alone when I feel it is appropriate to list the subject of my articles in abstract meta data. Interestingly, the number 2 result on Google for the search "enron scandal” (the BBC result above) does not include the word "enron” or "scandal” in the keywords metatag. The New York Times article did not even use the keywords meta tag.
More Resources and Links about Meta Tags
- Google Technology
* A potential exception is if 1) your meta tags include a unique or very rare phrase such as "cybertriallawyerscam”, 2) you use one of the few search engines that use keywords, and/or 3) you search for that exact phrase.