MUM — Answers that go further and beyond

Google’s MUM promises to answer you in a whole new way. But how do you usually search on Google? That depends, doesn’t it? After all, there is a reason behind every search. From a content producer’s point of view, the name of that is search intent. Each search intent generates a type of response. In other words, everyone can search for the same thing, but they do it differently.

And, above all, at different times. For the experts, this is linked to the moment you are in the user journey. That’s another subject, but let’s now try to understand why MUM can change the way Google will answer you without you having to change the way you ask. Or changing. And beyond.

Multitask Unified Model — That’s the name of your new private expert

Think you’re a user of a specific type of product. Let’s use a bicycle as an example. You use your bike to get to work, and you want to change it for a newer model. Maybe an electric model? Maybe. But you’re not sure. So, what would your searches be like? Bicycles for everyday use? Bicycles for going to work? Best bikes for everyday life? In short, it could be anything. After all, you are in doubt. Google and its artificial intelligence will realize this and adjust the answers to the moment it understands you are.

Meanwhile, good SEO and ADS professionals are wracking their brains to understand what that moment is and what keywords you would use for that search, and at what moment you are in the famous marketing funnel. After all, sending you the answer to a different moment than the one you’re at will have low or no effectiveness.

But that’s where MUM and its multitasking feature come in. What it does is break down the language barrier and the search stages. So we’re talking about reducing the number of questions you’ll need to ask to get to a more definitive answer. Yes, Google is paying to see and doubling down. Consequently, it wants to do most of the work for you. To that end, he wants to become your private expert.

“But as far as I understand, it’s more like a multi-purpose machine learning library. So it’s something that can be applied to many different parts of the search.

John Mueller, Google

So your query about bikes will already bring more information as it will compare many different scenarios involving “bikes used to go to work”. Your geolocation characteristics may be taken into consideration. After all, if you live in San Francisco, CA, full of slopes, your scenario is entirely different from living in Leipzig, Germany, much flatter.

Comparing things for you

Today if you want to change your bike and you are not a great expert, you will probably search for opinions. Also, will search for reviews, compare prices, payment methods. Maybe check about a delivery, accessories, and everything else related until the moment you finally decide to buy your new bike. Shortening the paths is what’s on the table with MUM.

Posting your current bike model will understand what you have and suggest similar, better, or more recommended ones. Logically, this will suffer interference from the excellent work of SEO and Google Ads professionals. But remember: Google prioritizes value delivery to the user. Whoever plays this game should be wanting to deliver value first and then seek to do business.

MUM has the potential to transform the way Google helps you with complex tasks. MUM uses the T5 text-to-text framework and is 1,000 times more powerful than BERT, its predecessor. In this way, it seeks a much deeper understanding than semantic understanding. It aims to understand what’s behind your question. If it cross-references information and realizes that you’re making comparisons, it can suggest things that maybe you were going to learn were important later.

Yes, it sounds daunting. But it’s at the heart of brands’ concerns. Getting you interested in complete packages. And Google knows how to target that.

And that’s not all. MUM also “reads” images.

Perhaps one of the significant revolutions is the way it reads images. If you post a picture of your current bike, it can understand that you’re looking for comparisons. If you add a search like “is it suitable for city use?” With that, get ready to receive suggestions for similar models, accessories, maybe even insurance, or even lists of best destinations and uses of your product or similar ones.

Who knows, you may even read reviews and indications of specialists and their cost benefits concerning other models. All with just a photo and a targeted search. In the coming months and years, we will be bringing features and enhancements to our products with MUM technology. While we are in the early days of exploring MUM, it is an essential milestone towards a future where Google can understand all the different ways people naturally communicate and interpret information.

Google Blog

Additionally, the language barrier is also being challenged. After all, trained in 75 different languages, MUM promises to give you value-added answers regardless of the language in which they were initially written.

So, it’s my friends. Nothing will ever be the same again. In fact, hasn’t this become the new normal?