Three Simple Relevance Tests

You have no way of knowing how search engines determine relevance but you can do some quick checks on a page to see what that page might be relevant to. Here are three examples of how you can test your page layout and content for relevance.

Relevance Test One – Circle The Key Blocks
You know you can divide a page into zones or blocks. Just print out a copy of your fully formatted Web page and circle every block of text (don’t worry about whether it’s indexable). Now create a list of keywords and add one expression (only one) for each circle on the page.

Each paragraph of text should be considered a separate block.

Each visible image caption on the page should be considered a separate block.

Each group of links should be considered a separate block (example: your nav menu would be a single block).

When you’ve finished writing your list, you’ll have an idea of how many topics your page is relevant to (don’t change anything because of this list).

Relevance Test Two – Name Those Links
Do you know how many documents/pages link to any given page/document on your site? Write up a list of all the linking pages that point to each page on your site (or a random selection). For each linking page, record its title tag and the anchor text it uses to link to your page.

Now divide your list of linking sources into two parts.

Part 1 includes all the links you yourself are responsible for (even if you only asked someone to place a link without specifying anchor text). This part includes all directory links, all reciprocal links, all paid links, etc.

Part 2 includes only the links you obtained naturally, without asking, exchanging, or paying for them.

Compare your two list halves to each other. What do they tell you about:

1. Your natural link profile versus your artificial link profile?
2. What other people think your page(s) is(are) about?
3. How well your opinion of your topics match other people’s opinions of your topics?

Relevance Test Three – If You Could Not Link To Yourself
You know who your competitors are but do they really have the most relevant content about your topics? Try to find ten documents that are more relevant to your targeted expression than either your content or your competitors’ content.

If you can build a list of ten such documents, ask yourself why you think they are more relevant than your own content to the topic. Is there any way you can match their relevance (on a human-intuitive level) by expanding or revising your copy (without infringing upon intellectual property rights)?

What would you do to those documents, preserving their content and structure, to help them rank in search results for their main topics? What could you do for them that you haven’t done for your own content?

The Takeaway
These three tests offer an intuitive approach to evaluating your optimization. You may be doing too much or too little of something. You may be overlooking something that is obvious to other people.

Ultimately, you want both your copy and your links to be natural and comfortable for your visitors. They should not be stumbling on glaring inconsistencies.

People, not just search engines, measure relevance. We measure relevance intuitively without conscious knowledge of algorithms. As SEO technicians we usually follow a set pattern of activities. Even non-formulaic SEOs are formulaic because you cannot keep doing what you’ve always done without being formulaic. The mechanics of our habits, formulas and procedures have an impact on the natural relevance we create.

Sometimes a document can be too relevant to a specific set of keywords and not relevant enough to a topic. In the same way links can be too relevant to keywords and not relevant enough to the topic. The more artificial a document’s connection to a search query appears to be, the less likely that document strikes the visitor as compelling. Artificiality diminishes acceptable relevance.