Loading Jaywing website
25 May 2016 / Opinion

Has Google become a semiotic search engine?

John Jackson / Head of Client Strategy

Recently, we’ve seen a change in the way Google operates. With an increasing focus on personalisation, are we really just looking at a semantic search engine, or has Google become a semiotic search engine?

What are semiotics and how does it relate to Google?

Semiotics analysis breaks down into three main sections, these being:

1. Syntactics (sign)

This is the form or layout of a word or phrase. Words, letters and punctuation in the wrong order can have a distinct effect on the user, causing them to slow, pause and mistrust the information in front of them.

For example, the phrase “dead is SEO” automatically sounds odd. The structure is wrong, meaning that as a reader you will place less trust on the source.

Syntactics were key for Google when it first started. Websites that were well put together and had accurate content were more likely to be useful for users.

Because of this, it was one of the key things SEOs first looked at. Web pages would be optimised for a specific term and that specific term had to be included a certain percentage of the time.

As Google grew, it started to make connections between plurals and synonyms which allowed us to expand the range of key terms included in our SEO content.

The importance of syntactics has lasted right up until now, although Google now uses a much larger range of factors when making judgements about your content. However, making sure that your copy makes sense is still vital for anyone looking to rank well.

2. Semantics (signifier)

The next stage in Google’s development was semantic search. Words can often have different meanings. Google (and people) try to work out the meanings by using other signifiers on the page.

So for example the term ‘alphabet’, can have two very distinct meanings. It can either be the company or a collection of letters.

A page on the company would contain things like:

  • Google
  • Business news
  • Stock prices

And less on the letter ‘z’…

Semantic search uses other signifiers on the page and the wider site in order to better ascertain the meaning and relevancy of the page.

Therefore, in order to rank well for terms like “drills”, the site you have should have some form of specialism in DIY, hardware or a profession that is likely to use that tool.

3. Pragmatics (signified)

Moving on a stage further from semantics are pragmatics. Pragmatics look at the fact that different words, colours and signs have different meanings depending on who you are, who you are talking to and in what context you are talking.

Below are a few examples of this in practice:

Example 1 - Consider searching for the phrase “who is the president”. The Pragmatics for the user will have a major impact. A search in France, for example would have a different answer to a search in USA.

Example 2 - In Roland Barthes 1957 piece, Mythologies, he points out how the signified can change depending on who you are talking to and what you are trying to portray.

Using red wine as an example, he states that the main signifiers point to red wine as being a red alcoholic drink that’s made of grapes.

Yet, the signified for the French Bourgeois was that red wine was a healthy, robust and relaxing drink. This is despite the fact that alcohol was known to be unhealthy and intoxicating.

The culture had placed a new meaning on the sign, signifying something specific to them and thus changing its meaning.

Wider cultures and audiences can do this with almost anything. Understanding how your audience understands certain terminology is vital if you are looking to engage with them directly.

Example 3 – As explained here, the term “you have a green light” can mean a number of different things depending on the context in which it is said:

  • It could mean that the space that belongs to you has green ambient lighting.
  • It could mean that you are driving through a green traffic signal.
  • It could mean that you no longer have to wait to continue driving.
  • It could mean that you are permitted to proceed in a non-driving context.
  • It could mean that your body has a green glow.
  • It could mean that you possess a light bulb that is tinted green.

Context is key for content. If the context is wrong, then the content can either not make sense, not be relevant and in some cases even be offensive.

Applying this to Google

Back in 2014, when semantic search was coming to fruition, a number of different SEOs postulated that we could be seeing Google move to a semantic model for search.

In an article for State of Digital, author Gianluca Fiorelli commented that:

“With the continuous effort Google is making in transitioning from an Internet of Strings (just based on the signifiers) to an Internet of Things (entities, both named and search ones, or signified in semantics jargon), we should soon need to talk of the passage from semantics to semiotics in search.”

A LunaMetrics article took a more technical approach, looking at the need for things like Schema mark up, the author wrote that:

“For ten years we have been doers, focused on tactics and finding cracks in the system to effectively game the system and get ahead. From now on we must be facilitators, providing clean, concise information to the search engine crawlers. Schema is one way to accomplish this. There will be others in the semantic revolution, like entities”

Moz author Isla McKetta also looked at how to apply semiotics back in 2014. She concluded with a number of interesting important ways in which pragmatics could be applied to things like web design.

Isla spotted back in 2014 that a move to a semiotic approach was likely and recent developments may be highlighting a definite move towards this.

Personalising the search experience

Syntactics is basically just good form for a website and semantics are now part of normal content development practice. The groundwork is therefore laid for the move to the next stage.

Personalisation is having a major impact on the search results. What you have searched before, your location, your language and your social profiles can all have an impact on the results you see.

We should also consider the impact of RankBrain. RankBrain gives Google the opportunity to learn how people search on a much greater scale than before. By using machine learning, it can now make associations between the user and the terms they use on mass.

Previous incarnations would always rely on linguists and quality experts, all of which would have their own biases. Assuming RankBrain works effectively, it should remove the biases and start to assign the correct meanings of words, depending on the context and the audience.

In Google’s quest to give users the best possible search experience, it makes sense that it’s going to focus on increasing personalisation. This means that making your site and the content on your site relevant to your users is more important than ever.