In the early days of Google, the search engine acted as an excellent alternative to traditional resources like dictionaries or the Yellow Pages (for those who remember that tome). It offered simple answers to easy questions, such as the definition of a word, the address of a business, or the time in a different part of the world.
As searchers grew to trust the tool, they began asking more of it. Over the years, Google continued to improve and refine its flagship product. Search has come a long way since its origins in the late ’90s. But with the engine fielding 3.5 billion daily searches — 15% of which have never been seen before — Vice President of Search Pandu Nayak says there’s still plenty of progress to be made.
If the incoming search algorithm update announcement is any indication, Google is taking the future seriously. The Multitask Unified Model (aka MUM) is said to be 1,000 times more powerful than its BERT predecessor. It aims to provide better answers to more complex questions, largely thanks to its natural language processing capabilities. Here are three ways MUM will change the future of search for digital marketers:
Better Understanding of Language Patterns
Nayak describes the current way people search on Google as “keyword-ese.” It’s a far cry from how we usually speak or ask questions — people would never walk up to their friends, say “best tacos near me,” and expect an exhaustive list in less than a second.
MUM represents a quantum leap forward in the power of a Google Search to comprehend what users are actually trying to say. As the algorithm’s understanding of language improves, there will no longer be a need for keyword-ese. Marketers can expect queries to look more like spoken language.
Enable More Comprehensive Query Targeting
MUM doesn’t just understand the language contained within queries. It can produce language itself. This capacity allows it to gather the most useful information from a wide variety of sources.
What does this mean for marketers? Even if a piece of content doesn’t provide an exhaustive answer to a query, it might still be useful if it authoritatively covers a more specific subsection of a topic. MUM can also understand and translate content produced in 75 languages, allowing searchers to tap into a global knowledge base.
Reinforce a Broad Content Marketing Strategy
High-quality content that engages users and meets their needs currently rules the roost on search engine results pages. This won’t necessarily change overnight; like BERT, MUM will direct searchers to a slew of useful written content. The difference is that it will also be able to take a query and identify relevant videos, infographics, podcasts, and other content types that add value to the user.
When a query can be answered through a variety of mediums, the possibilities for marketers are endless. And it’s no problem for marketers who have been focusing primarily on written content: This capability allows them to repurpose high-performing blog posts into other types of content.
By translating materials into other formats, marketers can tap into an entirely new audience and give existing fans and followers another way to interact with brands.
MUM likely won’t roll out all at once. The algorithm is still being trained to ensure it performs as predicted, but it has already left an indelible mark. To ensure the dissemination of trustworthy information about the efficacy and availability of COVID-19 vaccines around the world, Google used MUM to create a list of more than 800 keywords associated with the vaccine in more than 50 languages. The task might have taken human researchers weeks, but MUM accomplished it in a matter of seconds.