If you haven’t heard of the Google MUM update by now, then it will pay to get an understanding of what is to come, and how you can prepare your website for it. Let’s dive in!
What is Google MUM and how does it work?
Announced in May 2021 at the Google I/O virtual event followed by a blog post from Google’s Pandu Nayak, Vice President of Search, the Google MUM update (Multitask Unified Model) is an AI-powered algorithm that aims to solve one of the many problems users encounter with search: having to type out many queries and perform many searches to get the answer you need.
It intends to answer complex search queries and give more direct, human-like answers. The idea behind the MUM is that the AI within Google searches will analyse and generate content in the same way that a human does, and therefore answer a question, or search term, in the same way.
MUM is built on the same transformer architecture Google used to create BERT (Bidirectional Encoder Representations from Transformers), however it is expected to be 1000 times more powerful.
Here’s a quick summary of the key capabilities of the Google MUM update:
- Acquires deep world knowledge
- Collect and translate content from other languages
- Understand and generates content over 75+ languages
- Understand multiple modalities – text, images and video
- 1,000x more powerful than BERT
What are the capabilities of Google MUM?
According to Pandu Nayak, the AI model within the MUM update will be able to automatically and accurately detect a wider range of searches, including personal crisis searches (see below) and other search queries that require immediate answers.
This allows MUM to better understand the intent behind people’s questions in order to detect when a person is in need. This helps Google show more trustworthy, reliable and actionable information at the right time. (https://blog.google/products/search/using-ai-keep-google-search-safe/)
Below are some of the key capabilities of Google MUM:
Detect Personal Crisis Searches
Google recently announced that it started using MUM for better detection of queries around personal crisis. According to Google:
“Using our latest AI model, MUM, we can automatically and more accurately detect a wider range of personal crisis searches. MUM can better understand the intent behind people’s questions to detect when a person is in need, which helps us more reliably show trustworthy and actionable information at the right time.”
The MUM update essentially will allow Google Search to have a better understanding of what someone’s intention is behind a search and in the case of crisis situations, will provide appropriate results such as mental health helplines when typing in suicidal searches.
Google also started using BERT to reduce users’ chances of encountering unexpected shocking search results. This is most effective in filtering searches related to ethnicity, sexual orientation and gender. Google said “BERT has improved our understanding of whether searches are truly seeking out explicit content, helping us vastly reduce your chances of encountering surprising search results.”
Language Barrier Removal
The language barrier removal is one of the most exciting capabilities of the MUM update, and is a big improvement in search experience.
This means that you can search something in one language and receive the answer to your query, translated into another language. This is particularly useful for international queries or people wanting to find out information about a different language, culture or country.
Most of the recent Google’s algorithm updates (BERT, MUM, etc) are aimed to improve users’ search experience by understanding how they use search to provide more accurate results to a complex query.
One of the key capabilities of MUM is its ability to “learn” different languages. In short, it will allow Google to find information in one language and translate it into another to provide the most accurate information for the query.
For example, if you are planning to cook a particular Japanese recipe, there may be lots of “english” content on this subject but there might be more in-depth, authentic content in Japanese. MUM can take that information from Japanese and translate it to English or 75 other languages – providing users more accurate information.
Understanding information across types. For example, search for images and questions with Google Lens
Another exciting capability of MUM is multimodal which means that it is capable of understanding information across multiple formats (text, images and video). From the example that Google has given in their presentation, MUM is already able to understand information or content in both text and images. We expect it to learn content information from videos and audio in the future.
According to Nayak, “Eventually, you might be able to take a photo of your hiking boots and ask, “can I use these to hike Mt. Fuji?”
MUM will be able to understand the photo, take it and connect it to your question and answer it as accurately as possible. The search results can then link it to a blog, recommended list of equipment, etc.
How will the it affect my websites?
Essentially, MUM will ensure that more relevant content is prioritised over content that is less relevant with poor layout, and will disappear at a faster rate than with BERT. This means that for website content to rank, it’s even more important than ever that the user experience, overall SEO and accessibility are designed for quick and easy answers with the use of multimedia formats (images, FAQs etc) beyond the traditional realms of simply having the relevant keywords.
How can you prepare for the MUM update?
The end-users of a search query are the main focus of the MUM update and this must be at the forefront of content marketers and SEO minds when writing and optimising website content. It’s also important that web developers are aware of this as full websites and webpages need to be optimised for the end-users, whether that be on mobile or desktop devices.
Here are some of the ways in which you can prepare for the MUM update:
- Make sure all content is user-focused. All content writers must ensure that their content answers questions. In order to perform well and attract organic visitors to your website, user-focused questions must be answered within your content in order for your webpage(s) to appear in the SERPs.
- As mentioned previously, language barriers are being removed so you will be competing against web pages from across the world. Consider having multi-lingual friendly content that can be easily translated by Google. Avoid overly regional-specific language that can’t be translated by MUM.
- Continue to create high quality content but look into other content formats for a greater chance of being noticed by Google. Images, videos and other forms of multimedia will play a big part in the MUM update and will be used to help answer search queries.
- Make sure all images are fully optimised with the right file size, and include alt-text. Images and other forms of multimedia are going to play a big part in search results so it is important that they are of good quality and good SEO practice. It’s important to remember that the Core Web Vitals hasn’t gone away with the MUM update and user experience is very important to both.
- On-page information must always be optimised with the right keywords. This is a general SEO standard, but even more important with the MUM update. Ensure the keywords and in this case, the search query, that you wish to rank for are incorporated into your on-page information (page titles, meta descriptions and headers).
- Schema markups – Schema markups are also known as structured data which can be added to the HTML coding on your website. This helps a Google bot (or any other search engine bot) understand the keywords that you wish for your webpage to rank. Whilst this isn’t a new, exclusive change with the MUM update, it’s now more important than ever to ensure that the search engine understands the information your page can give in response to a search query.
Remember that we don’t know with 100% certainty how the MUM update will affect websites in the long run or exactly what we can do to ensure our website’s rank. All we know is what changes are expected to happen with the MUM update and use SEO knowledge to make informed decisions.
When will Google MUM start to take effect on search results?
As Google mentioned that they will start rolling out the MUM update over the next few months and years, there’s no certainty when it’s going to be 100% in place. Similar to BERT, Google will undergo a rigid process as it implements certain aspects of the update. They also mentioned that “we’ll look for patterns that may indicate bias in machine learning to avoid introducing bias into our systems.”
Nayak also mentioned the Google Search Quality Rater Guidelines – the human team of raters who assess the quality of search results it provides. This is where Google’s E-A-T (expertise, authoritativeness, and trustworthiness) guidelines will come into play.
While waiting for the full update to roll out, the only thing we can do to optimise and be future-proof ready for these updates is to follow basic Google guidelines. We will keep an eye on this…
FAQS on Google MUM Update
The Google MUM update, or Multitask Unified Model, is an AI-powered algorithm announced in May 2021. It aims to address complex search queries by providing more direct, human-like answers. Unlike previous algorithms like BERT (Bidirectional Encoder Representations from Transformers), MUM is built on the same transformer architecture but is expected to be 1000 times more powerful.
Google MUM is designed to automatically detect a broader range of searches, including personal crisis queries. Google aims to better understand user intent, especially during critical situations. The update helps Google provide more reliable and actionable information at the right time. Google is actively using MUM to enhance the detection of personal crisis searches, contributing to user safety.
One significant capability of Google MUM is the removal of language barriers. It can acquire deep knowledge in over 75 languages, translating content for more accurate results. This is particularly valuable for international searches, allowing users to receive answers in their preferred language. Content creators are encouraged to consider multilingual-friendly content to cater to a global audience.
To prepare for the Google MUM update, website owners should focus on user-focused content, consider language variations, leverage multimedia content, and optimise for on-page SEO. The update emphasises the importance of user experience, relevance, and accessibility. As for the timeline, Google plans to roll out MUM over the next few months and years, with an ongoing commitment to avoiding bias in machine learning during the implementation process. Stay updated with Google guidelines for the latest information.
Wriggle SEO Services
If you are interested in Wriggle’s SEO services, contact us for a no obligation initial consultation.
