Heard of the MUM update? Confused by what it means to YOUR tech business? MUM’s the latest word (ahem, we mean acronym) in the SEO-verse, and we’ll break it down for you today.
What is MUM?
Google’s Multitask Unified Model update—or MUM for short—is a new AI model that Google’s applying to its search engine. The first of its kind to be integrated into the Google search algorithm, MUM is set to transform the way we look for information online…and that’s a high bar coming from the long-reigning king of search.
How does MUM work for the end user?
With MUM, when you type something into the Google search bar—whether it be a query, question, comparison or anything else—the algorithm pulls information from text AND images, videos and other non-text content to populate the search engine results page (SERP).
It can also search using images alone or images with text as an additional modifier. To help you understand this concept, Google offers the following example:
Imagine you’re using the Google Lens app and you search for a photo of a person wearing a floral print t-shirt (fashionista!). You can then highlight that shirt and Google Lens will populate a SERP showing other shirts with similar patterns.
Cool, right? Well it doesn’t stop there. Let’s say that you love that floral pattern, but you wish you could find it on a pair of socks. 🧦 What you would do next is type in a query like, “socks with this pattern.” With MUM’s guidance, Google analyzes images and returns pictures of floral pattern socks.
If you wanted to, you could continue to add layers to your search and build on top of it as you go.
How does MUM work behind the scenes?
MUM works using the T5 text-to-text framework. It builds on 2019’s BERT update, a natural language processing (NLP) that generates search results based less on exact keywords and more on contextual, related content.
One of the defining features of the MUM update is its ability to break down text and language barriers to produce relevant, accurate and detailed search results.
MUM understands context and language in a more intuitive, human way, just like BERT does, but it’s 1000 times more powerful than BERT and it can understand and generate communication in 75 languages!
Also, MUM can perform countless different tasks at once in order to offer you more comprehensive and applicable search results than older models.
It can read your query in your language and find something relevant in, say, Japanese. THEN it can use the insights it gains from these sources to find the most relevant search results in your language.
What does MUM mean for the future of SEO?
All search engine algorithms up to this point have relied solely on text. For example, the YouTube algorithm looks at descriptions and video titles.
MUM is taking that next natural step, which is understanding what’s encased within all those non-text pieces of content just as well as it’s able to understand web page content.
So, will this change render SEO obsolete? We don’t think so!
Why? Because non-text content will have SEO value as well.
Many businesses are hesitant to create non-text content because non-text content hasn’t traditionally carried any SEO value. We can posit right now, that Google search results are composed only of information from text sources.
For instance, with the MUM update, videos can rank for certain topics even if they never explicitly mention said topics. This empowers content creators of all kinds to step outside the world of text and embrace other formats.
What should tech marketers do to adapt today?
As Google continues to roll out the update into the algorithm, we’ll continue to see positive change with Google Search over the coming months.
So, what should tech marketers do next? First, don’t stop creating and publishing that expert text-based content. Just because SEO is expanding doesn’t mean the value of text is going away!
Instead, take this as an opportunity to ADD different mediums of content to your marketing strategies. Explore different formats and see what works best for you.
We encourage you to start diversifying your content with videos, podcasts, and other forms of multimedia sooner rather than later because it will help increase your search presence as MUM grows.
Plus, publishing different forms of content across channels can boost engagement and results for your marketing efforts—it’s a win-win! 🤩
To infinity and beyond—what’s next in SEO?
MUM isn’t the final frontier of AI by any stretch of the imagination. So, what’s next? In the short term, Google is building new features for its SERPs enabled by MUM.
Google already features similar tools, like the “people also ask” drop-down menu, but this will be much more granular.
For example, if you Google’d “acrylic painting,” you could go to the “Things to know” drop-down and find more drop-downs relevant to acrylic painting, like:
- Step by step (How to start acrylic painting?)
- Techniques (What are fun techniques to try?)
- Using household items (What can you use around your house to paint?)
- Tips (Tips to get you started with painting)
- How to clean (How to get paint stains out?)
In addition to this very cool feature, Google will also be adding “refine search” and “broaden search” at the top of the SERP.
So, in a nutshell that’s it! MUM is about to hit the world by storm—what feature are YOU most excited for? 🧐