Any discussion on optimizing blog posts or webpages, would truly be remiss without a detailed explanation of the power of latent semantic indexing. Wikipedia defines latent semantic indexing (or LSI) as “an indexing and retrieval method that uses a mathematical technique called singular value decomposition (SVD) to identify patterns in the relationships between the terms and concepts contained in an unstructured collection of text.”
To put it more succinctly: LSI is based on the principle that terms used in the same context have a tendency to have similar or related meanings. As you might imagine, Google uses LSI as a way to measure, contrast and compare the relevance of your webpages (and website) to a particular set of inter-related keywords and other ranking factors.
When optimizing webpages, the traditional notion of ‘anchor text’ and ‘on-page keyword density’ can be very misleading to the uninitiated. What truly matters is not the number of times you strategically repeat your keyword, nor does it have anything to do with ‘hyper-linking’; it’s about the semantic relevance of a grouping of keywords to the overall contextual theme of your website in conjunction with other ranking factors.
You might be sitting there scratching your head saying, “Huh?” But ponder this altruism: it is possible for a webpage to rank for a specific keyword that is not included anywhere on the specific webpage, or included in it’s meta content. Yes, I’ll repeat myself, it is possible for a webpage to be semantically relevant to a keyword and rank for that keyword, and yet not contain that keyword anywhere in it’s on-page content. Don’t believe me?
Perform a Google search for the keyword “search optimization”. The third organic search result is this: http://searchengineland.com/guide/what-is-seo.
Now tell me if you see the phrase “search optimization” anywhere on that page or in it’s meta data…(you won’t). Case and point: Yes, the term ‘search engine optimization’ is sprinkled liberally throughout the page but not the term ‘search optimization’.
Ultimately, website SEO sculpting is more than simply sticking a few keywords in our title and meta description.
It all comes back to semantic diversity, website authority, website trust factors, contextual interlinking and overall keyword relevance.
Having experimented widely with many different keyword strategies in sculpting sites for search engine rankings, one thing I can assure you is that repeating the keyword you want to rank for over and over again throughout your website will only ensure that you never rank for that keyword. Semantic diversity is mission critical as more weight is given to a site’s overall central themes than to the specific on-page keyword densities sprinkled throughout the site. This is an important concept to understand when sculpting your on-page content.
When you perform a search on Google, the algorithm innately serves up pages from different websites, each being a slight variation on your search phrase, in an attempt to offer you a more relevant search experience. Think about this the next time you perform a Google search. Not all of the links presented are relevant to what you might be individually searching for, they are a variation of a central theme. To another user, what you consider a ‘variant’ or irrelevant result, might be actionable. Most Search Engine Results Pages (SERPs) are contextual matches.
Semantic diversity is key. But wait a second…
Now after this extended diatribe, I am not saying that you should exclude your primary keywords from your title, permalinks, meta data or H1 tags, I’m just saying you need to be very selective and strategic in making these choices. Though your main text content should include your main keyword or phrase, the content should never focus solely on that keyword or phrase. There is a possibility that Google may see your webpages as over-optimized, and trust me, it will become nearly impossible to rank for your intended keyword no matter how hard you try once you incur this penalty.
Here are some hard and fast rules:
- Don’t over-stuff your site with keywords! Keep your keyword densities to under 5%!
- It’s good to use your main keyword within the first paragraph so it can be used in search results, but please, use your target keyword sparingly.
- Don’t use the same keywords in your post or page titles, permalinks or H1 headings. Always try to use some variant semantically related keywords. Be selective and be strategic!
- You should, as a rule, try to use your main keywords in your meta title because this information is what is displayed in the search engine results.
- Every page and post on your website should have a unique title using the keywords that are relevant to those pages. Never stuff keywords into your page titles or meta titles that are not relevant to the on-page content.
- Because search engines cannot read images, use your keywords and key phrases in ALT attributes. These should be short, relevant phrases that describe the image. No numbers, no special characters.
- Anchor text: I have mixed emotions about anchor text in general. I’m no longer certain just how strong a ranking factor this is for the search engines anymore. I use it very sparingly, if at all, as I am very sensitive to how it is distributed across my website as a whole, but I will say that if you have a massive blog or you intend to blog everyday for the next 5 years, you will want to strategically deploy anchor text. Smaller websites with less content should not use anchor links. Reserve contextual links for use in your navigational structure.
Tools I use for semantic keyword discovery:
Rely on this one the most:
I also use these occasionally:
I also almost always review the keywords being used by the websites ranking for my target keywords, checking things like titles, on-page content and meta descriptions for the semantic terms that Google may have used when scoring that particular page. There again, I am very cautious and try to blaze my own path whenever possible.