You might have seen Google’s I/O 2023 event earlier in May, where lots of new and exciting advancements were announced with a huge focus on AI. This is a topic that has been front of mind for us at Vuzo for a while now (in case you missed it, check out our blog on How data driven AI helps get the most from your team and your budget!), and since ChatGPT exploded into our lives at the end of last year, it’s been no secret that Google’s accelerated its own desire to produce a more AI driven search engine.
The waitlist for Google’s AI offering, Bard, was removed and it was rolled out to over 180 countries and territories on I/O 2023 day. Whilst this is very exciting, there is still a very real need for the human checks as the system is not 100% foolproof – we asked Bard to write some copy for a mobility scooter/electric wheelchair client and it came up with copy using ‘take a stroll around the block’ and ‘get back on your feet’…!
So what is SGE?
Cathy Edwards, Vice President and General Manager, Search Discovery & Ecosystems, introduced the AI generated Search Generative Experience (SGE), and you can check the video out in full:
SGE is designed to bring generative AI into search. Edwards says it’s for “making sense of something complex with multiple angles to explore – when your question has questions”.
So what does that mean in reality? Well, the first example Edwards gives is if you were to search for “what’s better for a family with kids under 3 and a dog, Bryce Canyon or Arches”. So far, so not a usual search term, right? Well Edwards acknowledges that although this might be the information we want the overall answer to, it just isn’t how we actually search today. We’re used to searching ‘keywords’, that is one or two words at a time, or a short phrase with no stop words. It’s these ‘keywords’ that we as marketers optimise for SEO, or bid on in PPC.
Although PPC has the evolution via match types to operate around a variety of similar terms or searches with keyword matches, SEO has remained rather stuck in the mid 00s. For example, we still use search volume when conducting keyword research.
This announcement is not dissimilar to how Google announced Google MUM in May 2021. Its Multitask Unified Model powered by AI is designed to understand more complex queries, different tasks within a single search and function across multiple media sources in different languages.
When introducing MUM, the example used was looking for what you should do differently to prepare to hike Mt. Fuji after you’ve already hiked Mt. Adams. When asking a search engine, you’d ask what the different elevations were or the gear you’d need and many smaller questions, whereas if you asked an expert, you could simply ask ‘what should I do differently to prepare’. The overall idea here is that you are using search terms that more closely mirror human interaction, rather than picking keywords and piecing together information to get to your answer. Both MUM and SGE want us to interact with Google the way we would with a human and Google wants to return the information in the same way.
This isn’t all that new to those of use who use other forms of search. For example, we’re getting more and more used to voice search, between home devices, smartphone voice control and using it in our cars – when using this, we’re asking for searches with far more natural sentences than when we’re typing.
So what does SGE look like for users?
The searches using the new SGE approach produce what is called a ‘new integrated search results page’ – in reality, it looks like an Answer Box Result on steroids. The Answer Box Results (ABR) is that chunky box that appears above your search results, pulling a sentence or a paragraph to answer your query. For example:
In Edwards’ example for SGE, the query returns three chunky paragraphs in a top box. Sources have their own citation boxes to the right and there’s an option to expand, ask follow up questions or click some suggested follow ups.
So, where doesn’t this leave organic results? They’re now buried deeply below the fold, so when it comes to SGE answers, it looks like it’s an ABR or the frozen tundra for SEO. She goes onto say “So you can check out recommendations from Experts…and learn from authentic first hand Experiences” . SEOs will recognise two very important words in this sentence – Expert and Experience – key E.E.A.T values. Google has been saying for years your content must have Expertise, Authority and Trust with the introduction of Experience last year, and this is going to play a bigger role in content.
One of the most important things that came up in the announcement is that “People will always value the input of other people”. This backs up one of our favourite bits of advice – make the most of your customer reviews! Customer review content will be incredibly useful when potential customers are visiting your site, as will any question and answer content, such as an FAQs section. In order to ‘demonstrate knowledge’ about a topic (and adhere to those E.E.A.T. values), and to appear in these new Search Integrated Results pages, it will be important to know the answers to many of the questions customers ask about your product, and display them prominently on your site. Not to mention, of course, that customer reviews are a Trust signal in themselves, and so only help when it comes to SEO. Not sure where to start when it comes to analysing your customer review data? Vuzo has a tool for that – check out ReviewLab and book a demo!
And what about when you are further down the buying journey?
This all obviously makes sense if you’re looking for genuine information, and if you’re at the start of your buying journey and you’re still fact finding. But let’s say you already know you want to buy a pink velvet bar stool? Well, you’re now looking for eCommerce, but don’t worry, SGE is there to bury the organic results here too(!). It’s still going to present a much larger ABR, but this time it’s going to bring in products that match what you’re looking for, which will be coming from Google’s shopping graph. On top of this, sponsored PPC ads will sit above the SGE ABR.
This isn’t to say that the ABR will be a free-for-all. Liz Reid, VP of Search at Google says SGE will show answers for safer queries. In other words it’s staying clear of strict YMYL queries, that is “Your Money or Your Life” topics that could significantly impact readers’ health, financial stability, safety, or well-being. Common YMYL topics include finance, medicine, and current events.
And were there any other announcements to know about?
At the same time, Google announced it will roll out a new update for the helpful content system in the coming months. This will take the form of the introduction of a Perspectives filter, which will be appearing at the top of search results when the content returned might benefit from the experience of others. Google wants to source this information from “discussion boards, Q&A sites and social media platforms” – be prepared for a lot of results from Mumsnet and Reddit!
The aim is that search results will prioritise a “greater focus on content with unique expertise and experience.” – i.e. your generic content writer producing 500 words on how to create an accent wall might not cut it. Don’t forget as well that AI, MUM, and SGE are multimedia, so iStock images might not be good enough. When creating any content, you need to prove you know what you’re talking about and if a picture is worth a thousand words, your content is going to need to be extra special if you want those Authority listings at the top of the SGE results.
In short, amongst the gazillion announcements Google made at I/O 2023, there are some interesting developments that marketers need to be aware of, and adapt accordingly. AI is not going anywhere, and the focus on E.E.A.T. content that makes the search experience more useful for consumers is ever more important.