Should you include both singulars and plurals in your keyword selection process? The quick answer is yes. It would be ideal to say that your pages are optimized for both the singular and the plural forms of certain keywords, assuming your visitors are using both, but it can get confusing. Which should you use in any given circumstance? Where on the page should you put them? What about the keyword meta tag? Most search engines do treat singulars and plurals differently, so these are good questions to ask.
The important thing is to understand which one your customers are using or will use and to remember that, by mathematical necessity, only one can have prominence. Only one can go first in the title, for instance. So, it’s not as simple as killing two birds with one stone or covering both bases, because it is actually impossible to optimize for two terms equally. Now ask yourself if this is really what you want. Is a near fifty percent split on the same page for two popular keywords really desirable?
Attempting to cast a wider net is never desirable with a new site. If ten people a month search for “widgets” and ten more people a month search for “widget” and you optimize a page for both, do you then have a shot at all twenty? Perhaps, if there is very little competition for those words, but it is more likely that you have simply diluted your chances for both. You have cut your chances in half for each, and what’s worse you can’t take those halves and add them together because half of a losing position is still a losing position. Search engine results are listed in order of relevance. Google has only one first page. Exact matching is somewhat important. So, better to use one stone for each bird.
The ideal solution then is to use two pages, optimizing one page for the singular and one page for the plural, taking care not to duplicate content of course. If this is not possible for your particular situation, then you will have to choose one, and decide which one deserves prominence and how much prominence. Don’t forget that secondary keywords, derivatives of your root words, and related words, should be liberally sprinkled around the page in the less prominent positions: singulars, plurals, nouns, adjectives, verbs, synonyms, even antonyms.
Search engines have evolved in recent years to include semantically related keywords in their algorithms. Primary keywords in association with secondary or related keywords rank better than primary keywords alone. Because of this, it is possible to “over-optimize” a page by focusing too much on one keyword or phrase. Search engines use “latent semantic indexing” to examine clusters of related keywords in an attempt to rank pages more accurately and return the most useful results. So, include a good mix of secondary and related words in your page content, including singulars and plurals.
Keyword meta tags are different, however. Search engines have mostly abandoned this tag in favor of those related keywords that exist on the page proper. Not a bad idea. Hopefully you have enough information on the page to make the meta tag irrelevant. Most people agree that these meta tags are not very important, but that it doesn’t hurt to keep in the habit of using them. You can never know to what extent they are being used or will be used in the future. Do not repeat keywords in your meta tags unnecessarily. Relevance is more important here than number or density.
It makes sense to me that search engines would prefer to find secondary or related keywords in the content of the page rather than in the meta tag. Latent semantic indexing gives search engines the ability to group “Babe Ruth” and “baseball bat” together for the purposes of ranking. It does this by examining supporting keywords on your page and from other pages in the same category and by building a secondary index of these related keywords. It doesn’t need a meta tag embedded in the code to do this.
The importance and the extent to which search engines use latent semantic indexing is still somewhat of a mystery, but the fact that it exists can be demonstrated by using a tilde before any search term in your Google search bar (~keyword). You will see that searching for “~design” returns and highlights related terms like “designer,” “designed,” “designing,” even “architecture” and “construction.” There are plenty of instances where this doesn’t quite show up so well, and when it does it is not always on target: The #1 position for “~scoreboard” in Google search is held by a page that does not have the word “scoreboard” anywhere on the page or anywhere on the site. It is held by a page optimized for the word “results.” Google understands that “results” can be semantically related to “score” and “scoreboard” and so returns the highest ranking related term it can find. The page actually has nothing at all to do with sports or scoreboards (results.org is a political organization), but the point is that Google does use some form of secondary indexing of related terms, with or without the tilde. A search for “widget” returns “widgets” in the #1 position. Most search engines also now include links to “related searches” right in their results pages.
Let me repeat: use both singular and plural forms of your keywords and related keywords liberally throughout the page and decide which one should have the prominent positions (do not try to split the prominent positions equally). You will get credit for your secondary keywords, by themselves and as support for your primary keywords. Repeat them often on your page, but for meta tags it is not necessary to include all forms or to repeat them because it’s little help, and search engines can parse these terms for whatever they’re worth: “scoreboards” already includes “scoreboard” within it and therefore using both is redundant. You will get equal credit for each. The same holds true for phrases which include other terms or phrases within them. Example: wood, wood stove, wood stove cleaners. The first two are redundant because they are contained within the third. Only the third one is necessary. “Wood stove cleaners” covers “wood”, “stove”, “wood stove,” and “wood stove cleaner” and “wood stove cleaners” all in one three word phrase. That’s five search terms in one which covers a lot of area. It is not necessary to keep repeating words in your meta tags. Key phrases are slightly different than keywords, as when you decide to include “wood stove pipes” in the example above, but you do NOT get a keyword density score (for good or bad) from your meta tag. It is doubtful that you will get much of a score at all, except where page content is missing or terribly light. And then not much. Keyword tags are not meant for people (they are meant for search engines) and are not considered content. The practice of stuffing these tags with irrelevant (or otherwise missing) keywords made these tags untrustworthy.
In conclusion, Google certainly does treat singulars and plurals differently, but it also considers them semantically related. This is why a search on either phrase returns mixed results. Google purportedly uses over 200 criteria for ranking pages and “exact matching” is just one of those things. Optimizing a page for the singular does not mean it will place higher in the SERPS than a page that was optimized for the plural when there are so many other factors for Google to consider. But it’s all about context and relevancy (and semantics). So, treat each case separately, give one prominence over the other rather than trying to use both equally, don’t repeat words unnecessarily in your meta tags, and then cross one more thing off your list of two hundred things to consider. :))