Machine learning and artificial intelligence are already reshaping the technological landscape. What was once science fiction is now becoming reality, from digital assistants and image-recognition software to self-driving vehicles. But what does this mean for marketing and advertising executives?
It might help us get closer to one of advertising's most coveted goals: scaled relevance. We will see changes in the way we do business before then.
Learn about sales enablement.
Technological advancements have always opened up new avenues for storytelling and marketing. Just as the introduction of television heralded a new age of genuine mass advertising and reach, and the internet and mobile devices ushered in new degrees of targeting and context, artificial intelligence (AI) will transform how people interact with information, technology, businesses, and services.
Google's goal statement is "to organize the world's knowledge and make it widely accessible and valuable" - and they've done just that over time. Google has created algorithms to better understand how people seek and organize information throughout the world.
In a nutshell, Google is attempting to train its robots to think like any human being, independent of language. This blog will focus on Google's two most recent developments, and how they have changed the way paid search accounts are established and handled.
Google AI Advancements: The MUM Update
Google’s new Multitask Unified Model (MUM) update, which was announced back in May 2021, was the most significant change made in the last 12 months and is believed to be a thousand times more powerful than BERT. But what is BERT?
What was the BERT Update?
The Bidirectional Encoder Representations from Transformers (BERT) upgrade, announced in 2018, improved the way Google AI matched queries with search results. This model was built on “transformers,” which are models that process words in relation to other words inside the sentence rather than independently and in sequence. BERT models evaluate the entire context of a word by examining the words that come before and after it, which is very beneficial for understanding the purpose behind search requests.
The primary consequence of this update was the addition of close match variations to exact match keywords, beginning with plurals and misspellings and progressing to other words and function words. In 2018, Google also added keyword variants, such as inferred words and paraphrases, that have the same meaning as the term. In short, “exact” no longer always means “exact.”
But back to Mum…
MUM is a watershed moment in AI since it not only understands but also produces language. Because it has been taught in 75 different languages, everyone should profit from this breakthrough.
Google recently showcased MUM's powers by presenting over 800 vaccine names from around the world in only a few seconds. Normally, this procedure would take weeks to complete.
MUM is also multimodal, which means it can interpret information in both text and pictures. The model will soon be able to understand a website's copy as well as its pictures, allowing MUM to better grasp the context of a page and deliver better results to the user. Google aims to develop MUM in the near future so that it can comprehend audio and video formats as well.
The Impact of MUM on Paid Search
Broad match modifiers and “clever” broad match are being phased out.
The Phrase Match behaviour upgrade began rolling out in June 2021 and was completely accessible to all languages the following month. Broad Match Modifiers (BMM) enabled marketers to broaden the reach of keywords without going overboard, giving them some control over the queries that were generated.
However, the most recent Phrase Match upgrade combines the enhanced reach of BMM while still offering advertisers control over searches.
Image Source: Google
Marketers have not been able to create new BMM keywords since July 2021; instead, these keywords now function like Phrase Match keywords. In summary, you can only run three distinct match types in an account (formerly four), a change that is simplifying keyword management.
Furthermore, Google just announced another change aimed at improving the accuracy of query matching for Broad Match terms. Broad Match keywords used to play a significant role in driving impressions and clicks, but they were frequently responsible for initiating inappropriate inquiries, so marketers had to carefully filter them out to keep traffic relevant.
The introduction of MUM and its capacity to interpret photos and text is unquestionably a significant improvement, since it now has a much greater knowledge of the context. The multimodal component of MUM is critical to the improvement of Google AI's query matching process. We wouldn't be shocked if a new "Smart Matching" match type became available in the near future.
By limiting paid search marketers to just one match type instead of two or three, this new “Smart Matching” match type would radically transform the way they structure their accounts. This, in our opinion, would be a substantial improvement because it would allow for a less complicated structure, easier keyword planning, and more complete ad-copy creation.
This move would also improve machine learning and bidding automation since traffic from relevant inquiries might focus on one term instead of two, allowing the system to learn rapidly from more data.
What’s Next?
AI advancements will keep people returning to Google, especially as search gets more conversational and machines continue to develop an incredible grasp of user intent, the human brain, and many languages. However, keep in mind that algorithms only see what they are fed, and human involvement will continue to be essential in conducting sponsored search activity.
Understanding a company's goal and converting it into a paid search strategy will continue to be a critical component of pay-per-click marketing that the machine cannot do (yet).
The MUM upgrade has just recently been released, and we have yet to see all of its features and benefits. But if there is one thing we are certain of, it is that the approach should allow marketers to spend more time creating engaging content, cultivating accounts, and exploring for new income possibilities, rather than focusing on manual, lower-value tasks.
As a first step, we propose that marketers evaluate their existing strategy. Then, companies should devise a strategy for streamlining their accounts, such as simplifying keyword management and utilising broader match types to bring in more traffic that may not have been collected earlier.
If you’re interested in more information about how to optimise your Google Ads, then contact us today.
