Google’s Multitask Unified Model efforts to deliver better quality content in search results are getting better. Thanks to AI algorithms that cannot only understand the sentiment in a text but also deliver search results based on content written in a different language than the one the searcher is using.
This is where MUM comes in. According to Google, MUM stands for Multitask Unified Model and will be a new milestone for understanding information.
Because this news is fresh and Google shared it with us yesterday, May 18, 2021. There is little we can say about it, but one thing I can glean from the release is that the problem with people plagiarizing content written in other languages is about to be medicalized.
How powerful is the Multitask Unified Model – MUM?
“MUM AI is like Bert” but 1,000 times better. And Bert is certainly a strong algorithm. So strong, in fact, that Bert was renamed DeepRank after a series of updates.
All this effort by Google to understand faster and better from content pages means that SEO will change, forever.
If you don’t keep up with these updates, you may find yourself doing something good today, in 2021, but there on 2022. Yes, it may be good today to rank better, but not tomorrow.
So to keep up with updates, stop worrying about keyword density. Instead, focus on topic relevance.
With the MUM update, if it comes, I can predict a severe loss of organic traffic for so many sites on the internet simply because they focus on covering a keyword and not highlighting a topic.
Here is graphical example of how MUM works.
10 Improvements Google’s Mum Will Bring to Google Search Engine:
- Answers to more complex and lengthy queries.
- A better understanding of the language to provide helpful answers.
- More complete answers to more complex queries.
- Completing broader queries with fewer searches.
- MUM can understand the search intent of a keyword more deeply.
- MUM can reveal insights based on deep knowledge for deeper exploration.
- MUM is capable of understanding information in multiple languages.
- It can also understand article formats.
- It pays special attention to semantic patterns that indicate machine learning biases.
- It is built on a transformer architecture.
You don’t want to be behind the news. Check back with us regularly to stay up to date on the latest Google updates, or contact us if you’d like to check how prepared your site is for these updates.