Google is changing, forever. I know, it sounds like clickbait, but it is a reality that businesses need to start paying attention to.
It is a little known fact that Google updates their search algorithm multiple times per day, in order to optimise millions of data points, providing you with the most relevant results. However, once every few years, things are shaken up drastically with the announcement of paradigm altering updates.
The most recent change is the introduction of Google MUM (Multitask Unified Model). It places an emphasis on search intention and a large amount of other data points; using AI & Machine Learning to give your customers the single most relevant and appropriate result, rather than a list of answers.
Google MUM
Imagine, if you will, how you would ask an industry expert a question. It would probably be structured similarly to this- “here is my current situation, some context on what I have done to date, here is what I want to achieve and what is the best solution based on the parameters I have specified.” The expert would also take what you have said in order to fill in the gaps with their specialised knowledge and provide you with one curated answer. This, in a nutshell, is Google MUM; a smarter and more personalised algorithm.
For example:
Please Note: This is simplified to make this infographic more accessible to a wider audience.
Compare this to the experience of using a search engine (currently) which goes along the lines of the following (as you can see it requires multiple searches and a lot of manual mental computations).
The new Multitask Unified Model, in essence, is moving one step closer to being the virtual assistant we have seen in science fiction…
What else can MUM do?
- Analyse results across multiple languages
- Analyse images, audio and video in the same way BERT (the previous algorithm) did with text.
- Skim read all data to better understand the contents of each page.
- Multi-Modal Search: combine both images and text into advanced queries.
- For example, While having lunch in Britomart, yousee someone wearing a trendy pair of shoes. You could sneak a photo and ask Google where they can be purchased from. Results would then pop with the product at a local store on queen street along with directions and prices for the item.
- Natural Language Processing with their T5 (Text-To-Text Transfer Transformer) AI model to convey intent and provide the search engine with the same level of comprehension / ability to break down text as a human would in a conversation.
https://ai.googleblog.com/2020/02/exploring-transfer-learning-with-t5.html
But what does all this mean practically for your business?
#1 – Focus on the desired journey & intent
In the world of advertising, we often visualise things like customer personas, UX journeys, campaign funnels, etc. Through my research into Google MUM, I have surmised that “SEO Intent Mapping” (a term we are coining if it doesn’t already exist) will become a more mainstream practice going forward. This journey mapping will coordinate with a revised keyword strategy (mentioned further down in this post) to focus on intent and desired outcome.
Practically, when finding a Long-Tail Keyword that is a low-hanging fruit, ask yourself four fundamental questions:
- What is the searcher actually wanting to achieve (intent)
- Where are these searchers (just starting, ready to “convert”, curious, etc) in the journey to completing an action (ie. buying a product, baking a cake, selling their car, going on holiday)
- What information is going to be helpful to the searcher (and let’s be honest, helpful to you as a business)
- How will searchers want to analyse this information
- - Full Article?
- - Snippets in Google?
- - Answer with some wider context.
- - Video
- - Infographic
I have prepared a Google Sheet which you can copy and use to help map this out for your business.
To learn more about intent, read Google’s Search Quality Evaluator Guidelines
#2 – Reimagine Blog Posts & Content
MUM is definitely a paradigm shift - transitioning from text, to multiple formats of data. Previously, every SEO specialist would tell you “you need to create more blog posts using a plethora of keywords”, but what we are learning is that we need to reimagine what we are creating to truly excel in these uncharted waters. We need to speak directly to the readers state-of-mind and deliver this in a way that balances Google's preferences with what best communicates the intended message that is rich & engaging.
When creating your next piece of content for your website, try applying the following criteria in order to maximise the effectiveness of your content.
Content Checklist
Before Creation
- What Long-Tail Keywords are you going to focus on?
- What keywords are you going to include?
- Who is the intended reader of this piece?
- What value are they trying to gain from this page / article?
- Where are they in their journey?
- How can you utilise different forms of media?
- How can you incorporate structured data into your posts?
While Creating
- Consider if your key points will make sense if read as a standalone excerpt. For example, while read aloud via Google’s voice assistant or as visual previews in search engine results.
- Is there any imagery that you can incorporate to emphasise a point (ie an infographic and/or an image containing unique identifiers that can be analysed by Google
This Checklist is also available in our template you can copy - Here
When you think of the intent behind this change in algorithm, it is clear that Google intends to make voice searches (Google Home, Android & their iOS App) a priority. Realistically, it will be more effective to have a tailored and personalised result in comparison to an over-saturated list of a seemingly infinite list of generic results.
#3 – Utilise Structured Data (Schema)
Those that know me in the context of the world of SEO know that I could talk an ear off a brass monkey about structured data, with good reason. Through the strategic use of structured data (also known as Schema), you can make optimal use of the featured snippets to better source organic traffic. The algorithm has built on and improved the capability of complex data analysis and as a result, featured snippets are likely to be more visible, tailored and widely adopted. Utilisation of Schema on your website, allows you to use the search algorithm to your advantage and make the best of this prime real estate.
Google MUM has the capability to analyse and summarise complex content to present data in a structured format. Google MUM has the capability to better analyse data and present a more personalised and text
This not only makes Google's job easier, but also allows Google to consider your content to be used in featured snippets which appear at the top of the page or make your result stand out from the rest.
-Featured snippets more visible and specific and common
-Utilise structured data on website to ensure you can get the most out of this prime real estate
Featured Snippets look like:
And to learn more about the different types of structured data, read this page by Google.
#4 – Go deep first, then, go wide
Traditionally, on top of the technical aspects of SEO lies a desire to fill your site with as many keywords related to your business as possible, distributing them between landing pages, information pages, blogs & other locations. But as Google transitions into utilising their Multitasking & Unified approach to allow more complex queries, the approach will need to morph its focus deeper towards prioritised terms (based on intent) before going wide.
Remember the packed query we were talking about earlier – “here is my current situation, some context on what I have done to date, here is what I want to achieve, what is the best solution based on the parameters I have specified.” Well, there is a lot to unpack, and right now, Google will take snippets (short-tail keywords) and display a collection of disjointed results.
I tried an example of this while searching for a new TV for my living room – “What TV should I buy between $1200 and $1800 that has Apple TV enabled and has dolby atmos built in”. The subsequent results were not entirely accurate. Google analysed my search content in snippets, resulting in the display of an Apple TV unit as opposed to a TV with the above mentioned built-in features. Furthermore, the shopping tab was equally non-specific- displaying normal TVs that did not match the pricing and feature parameters that I used in my query.
As Google transitions to supporting more complex queries, we (as searchers) will start to take advantage of this in our everyday life, and then we (as marketing professionals or business people) will need to cater to deeper searches with deeper keyword strategies that account for a vast array of complex intents, contexts and personalisation to continue reaching top-pole search positions. Currently, we can get away with single-word keywords and let Google put us in front of customers for a wide variety of search terms. Now with the complex algebraic-like calculations, you will need to rely on our deep understanding of customer behaviour and SEO Intent Mapping to ensure you don’t miss out on organic traffic going forward.
Timeline & Closing Thoughts:
Google is being very cagey on the release schedule, but small snippets are already starting to roll out, with small pieces here and there. So, in the meantime, we need to be preparing for this transition by leaning our efforts into embracing this upcoming paradigm shift and creating a rock solid strategy.
Actions from here:
- Find three use cases for Multi-Modal Search in your business.
- Identify the ratio of long-tail vs short-tail keywords in your strategy.
- Create a plan of how you can start incorporating intent into your strategy.
- Identify five ways you can use structured data on your site.
