Google Search has always been about providing users with the most useful results. 2019’s BERT update is a testament to that. You may think there’s nothing more to be done. Well, Google Fellow and VP of Search Pandu Nayak beg to differ. According to him, there is always room for improvement. The internet has been speculating about the use of artificial intelligence in search engines and it’s going to be real soon, with MUM, the latest Google algorithm update.
What is MUM?
Built on an all-new technology called Multitask Unified Model, the MUM update aims at providing results to users without having to make many queries and perform a large number searches to get the answer they seek. Here at Ranking Monks, we think this acronym is not just a happy accident. Because with this update, asking Google will be like asking your mum! MUM update lets Google to understand the nuances hidden in your search, especially when it involves the comparison of two things.
Consider this. If you’re preparing for a trip on another continent, you might want to do different searches to know about how the climate varies from your current location, the medication you’ll probably want to take, the clothes and accessories that’s apt for the climate and culture, the travel expense, hotels, and many more. Imagine how many searches you need to initiate. According to Google, users do eight queries on average to find answers for complex tasks. With the new update, Google hopes to resolve everything in a single search query.
BERT vs MUM
Bidirectional Encoder Representations from Transformers (BERT), a neural network based technology, was one of the major Google algorithm update made almost two years back. It allows Google to understand the search intent, rather than just processing the keywords and displaying some web pages that have keyword matches. MUM is also built on such a transformer architecture but a ton more powerful. MUM is trained across 75 different languages, elevating its world knowledge and comprehension of information.
MUM can not only understand your search intent, but also understand the context in which you’re making the query. This enables Google to provide additional information related to the context. MUM can also transcend the language barriers since it’s trained in several languages. If you’re planning to visit a place in Germany, it can now show you results from a German website (you probably won’t find results from other languages today) that would be more relevant to you than a website in English.
Since MUM is multimodal, it can also understand information from different formats like images and text at the same time. For example, you can take a picture of a lawn mower and ask “can I use this to cut the grass off my backyard?”. MUM would understand the type of lawn mower from your image you used and let you know if it’s appropriate. You might also get additional information on other accessories you might need for lawn mowing.
How MUM will affect SEO
MUM is still a work in progress and there is no official confirmation about the release. But for sure, it will have everyone’s head spinning. But as usual, SEO experts always find a way around. Judging from what MUM is planning to offer its users, websites will have to offer high-value content. It has to be comprehensive for sure, but will be better if comparison studies are added. On the bright side, you don’t have to worry about the language anymore. So, a French restaurant won’t have to translate its content to English to be ranked to be shown on the result pages. Better optimization of images, video, and audio, other than just adding alt text, will become mandatory. Finally, voice search optimization will be more important than ever.
MUM is still in its early years to be recognized as a major Google algorithm update, but it’s inevitable. Google wows to use the immense power of AI responsibly with the help of a large team of human raters. It’s always challenging to monitor a constantly evolving technology like Google Search and update your SEO works based on that. But hey, that’s the fun part, right?