The Body Text Is the Bread and Butter of The Page

The Body Text of The Page

The text on your site is its content in the eyes of Google. The text is the site. This may not be one hundred percent true anymore, as search engines have become somewhat better at interpreting and understanding other content too, though nowhere near as good as they are at interpreting text.

Text is often an area where sites tend to fail. Many publishing systems handle many of the technical details well enough to be able to take at least some positions. However, when it comes to text, it is up to you, and this is almost always very time-consuming. In the earlier days of search engines, the amount of text was an important factor for how well a page placed in the results. The number of pages was also important; more pages meant more opportunities to rank on more keywords. Strangely enough, this still has a strong correlation with how positioning seems to work today. More tends to be better, with some exceptions.

We have run a long series of tests by simply increasing the amount of content on a specific page and then measuring how it moved up in the results. In the past six months, the page that has managed to climb the furthest in the search results for ‘Search Engine Optimization’ is soekmotoroptimering.se, which is run by Håkan Persson. At the moment, this page has no fewer than 5,553 words. It’s no coincidence that he has chosen to put such a large amount of text on that page.

Text is of course not just text. The most common words used about text that is good for search engines are ‘relevant’ and ‘detailed’. Simply rambling, or adding nonsense text, is something that can actually damage your search engine optimization. You need text that is packed with good information for your visitors and which is relevant to the keywords that you want to be visible on. This means, amongst other things, that you need to mention the exact keyword you want the page to be visible on. But beware of stuffing the text with that word.

Our general advice is to write text that is as good as possible for the user and contains as much information as possible. Feel free to add more text, but do not add unnecessary filler. Make sure that the text adds value.

Google is happy to talk about how good they are at understanding text; and yes, the search engine is good at recognizing synonyms and understanding context. However, they are way off actually understanding the text. Therefore, you need to be clear about which keywords the page should appear on. It is not enough to hint at the keyword; you need to be clear and spell it out exactly. If you want the page to rank on ‘Really good screws’, then you need to make sure to add ‘Really good screws’ to your text. While it definitely doesn’t hurt to add ‘Very nice screws’ or ‘Really good bolts’ to the text, the most important thing is to remember the main keyword.

Another important aspect is that search engines find it difficult to accept that one page can be about more than one thing. We have already mentioned this in the section about landing pages, but it is important, so it doesn’t hurt to clarify it. For maximum effect, each page needs to focus on only one thing. It is often a matter of setting boundaries.

In the past, Google couldn’t handle the fact that the plural and singular versions of a word were still the same word. It was better to make one page about Nuts another one about a Nut; at least if you wanted to maintain good positions on both keywords. The same applied to word order. Games online and online games were completely different in Google’s eyes. It is somewhat better now, and our usual recommendation is not to allocate these things to different pages, but to try as far as possible to rank on, for example, Nut and Nuts with the same page. It does not always work, but it is a good starting point. Another aspect that needs to be taken into consideration here is Panda; but more about that later.

Re-using the example from the landing page section, we have three different pages; one about bicycles, one about mountain bikes, and one about Scott Spark 900 SL (which is a mountain bike). This is a pretty reasonable and common division for a web shop. The home page should rank on the keywords Bicycle and Bicycles, while a category page should be visible on the keywords Mountain bike and Mountain bikes. At the bottom of the hierarchy, we find the product pages that should be visible on the name of the products and any variations of these, in this case Scott Spark 900 SL.

It can be difficult to get a good understanding of what the actual division looks like when it comes to the text. The first page should be about bicycles. It should not be about mountain bikes or about Scott Spark 900 SL. However, it is not a bad thing to include references to various types of bicycles, including mountain bikes. To clarify, we will add some examples of things to discuss in the text below:

Different types of bikes

What to think about when buying a bike

The history of the bicycle

What differentiates our bikes from the competition?

What distinguishes this page about bikes from one about mountain bikes? You can call it semantics, but simply switching the topic from bike to mountain bike will make it a page about mountain bikes.

Different types of mountain bikes

What to think about when buying a mountain bike

The history of the mountain bike

What differentiates our mountain bikes from the competition?

This will generate completely different texts. There are different things to consider when buying a mountain bike compared to when you’re buying a generic bike. The mountain bike has a completely different history, and so on. There is nothing strange in mountain biking being mentioned on a page about bikes, or bikes being mentioned on a page about mountain biking. The important thing is that the text is relevant to what you want the page to rank on. The goal should be to create good and natural text that deals with that particular subject.

When it comes to the individual products, there are some things that are worth keeping in mind. A product page usually contains text that comes with a particular type of content. This is usually text describing of the product, as well as product information and the price. Many e-retailers choose to re-use the information they receive from the supplier, and when it comes to product data, there is not much that can be done about it; the bike weighs what the bike weighs. However, regarding the body text, re-using text is something that should be avoided, and we will explain why shortly.

Product pages, or the pages at the bottom of a site’s hierarchy, usually create the most problems when it comes to text. An e-retailer that sells around 10,000 products does not particularly like to write 10,000 different pieces of text, especially not with the motto ‘the more detailed, the better’. It is obvious that this is a massively onerous job. The benefits of implementation are great, but, as always, you need to weigh the pros and cons. In the online shops that I own, we choose to publish products with the suppliers’ texts at the outset, and then rewrite them over time. It takes time, but we find that it is worth the effort.

To get back to why it is bad to use the suppliers’ text, the reason is simple: your text needs to be unique for Google to consider it valuable. No search engine really wants to display several versions of the same text in a search result, so an algorithm has been deployed to ensure that this will not happen. Accordingly, copies of texts (called Duplicate Content by Google) suffer in search results. Duplicate Content is filtered, and the search engine tries to find the original source and only display that one.

In the course of the last year, one piece of text that has bounced around a lot in the search results is, hilariously enough, Google’s own Basic Manual on search engine optimization. This usually maintains a really good position on important keywords in search engine optimization, but over time, a copy of the same text (one of many) has taken the position of the original. Due to the Duplicate Content algorithm being far from perfect, Google has periodically lost its own version of the text and then chosen to display someone else’s copy instead.

So there are good reasons to make sure that the text you have on your site is unique. In fact, it is not even enough that you haven’t copied it; it is also important that no one else copies your text, as you can see from the example of Google’s own Basic Manual.

A common question that pops up when people first hear about Duplicate Content is how many words need to be replaced in order for the text to be counted as unique. It is perhaps both a blessing and a curse that Duplicate Content is not binary; it is not on or off. The uniqueness of a text is on a scale. You can’t replace 10 % of the words and then expect Google suddenly to think that the text is unique. If you change 10 % of the words, Google will consider the text to be 10 % unique. This means that it won’t necessarily be harmful to quote someone, for example, but neither can you simply search and replace text in order to make it work optimally. The text on all your pages needs to be unique. This usually requires hard work, but this is what is needed if you want to do it properly.

There are advisors in search engine optimization who have said that every time you mention a product or service on the landing page, you need to match the exact keyword. This is simply not true. Google gets better and better at managing synonyms and understanding the context in of a text. In fact, using synonyms, plural forms and related words is something that actually strengthens the page. It is important that the keyword is mentioned, but it doesn’t hurt to use words in the same ‘word cloud’ as well.

There is a theory that Google uses Latent Dirichlet allocation, a model that can tell if there are similarities between two texts. By using this or a similar algorithm, it is possible in theory to see the difference, for example, between an article about the Iron Man movie and one about the Iron Man competition. You simply compare and group the information with other information in the word cloud.

There is currently no evidence that Google uses such an algorithm, but an educated guess would suggest that something similar is in play. Synonyms and putting your text into context seem to produce good search results.

Sources:

http://static.googleusercontent.com/media/www.google.se/sv/se/intl/sv/webmasters/docs/search-engine-optimization-starter-guide-sv.pdf

https://support.google.com/webmasters/answer/66359?hl=en

http://moz.com/blog/lda-and-googles-rankings-well-correlated

http://en.wikipedia.org/wiki/Latent_Dirichlet_allocation

Next Step

All pages need static content