Google’s annual I/O event got off to an exciting start with its keynote. In summary: AI is now ubiquitous in all of its core products. I’m going to cover the ways in which it will most affect SEOs and content marketers in coming months, but also touch on the impact for developers and creatives as well.
If you haven’t yet watched the Google I/O keynote from today, and you want to, you can find it here. You’ll want to skip the first hour or so of the countdown before the actual keynotes begin.
Staying on top of it all
AI (and its impact on SEO) is evolving fast enough that the pace is starting to feel frantic.
There is a very real sense of anxiety that I get when I talk to my peers and it boils down to: how do I stay on top of it all?
The key to staying on top of it all is having a high enough level of abstraction.
In the context of Google search and anything related to content and how it will be affected by AI, the easiest way to think about is this:
Things are going to get wild.
Today’s keynote was the first time I’ve seen Google, in quite a long time, narrate a coherent vision for where their products are going, especially in light of the AI revolution.
They did a good job of identifying the core problem that users are trying to solve in each of their major products, including Google Search, Maps, Photos, Workspace, Mobile, and more.
Impacts on Google Search
Long story short, Google should probably rename their blog from:
Their post on “Supercharging Search with Generative AI” describes the major changes that are coming.
First up is the rolling out of generative AI as (essentially) a SERP Feature in the search results.
This is a change from the user’s perspective (if it works well) because instead of each individual search being a discrete event (manifested as keywords today), we’re moving toward a search being a prompt that generates a new context, which will end up being a much longer engagement for both the user and for Google.
The further integration of Google’s Shopping Graph into this feature is where things will get very interesting for eCommerce marketers.
Marketers will still have the option to list products in the Merchant Center or ensure that they are surfaced via metadata on their own sites (or both).
What will be enhanced, again from the user’s perspective, is much more navigation through product categories via criteria that are specified in prompts. This means that merchants will need to ensure that their product lists and metadata are very well-optimized and discoverable.
There is clearly a lot more coming but in terms of immediate SEO impact to plan for, I’d recommend starting on these topics with your team.
Content creation is totally disrupted
A few months ago, we made the following observation:
And we stand by this. After today’s keynote, even more so.
There is just so much useful innovation coming in Google’s core products around this that will be difficult for users and content creators to ignore.
Google’s new “Help me write” feature will write documents, emails, job descriptions, and more with a simple prompt. You’ll be able to refine prompts and expand the content with the click of a button.
The two areas where we see automated writing tools for SEO purposes having an advantage will be 1) scoring content and 2) generating content that is not watermarked by Google as synthetic content. How much this matters over time remains to be seen.
Content creation agencies will need to adopt much more of an engineering focus going forward.
Aside from the updates to Google Search, the Workspace upgrades were most impressive.
Image editing and generation tools
There is a whole new range of image editing tools available in Google Photos, Google Slides, and more.
Creatives will either love or hate these features. Or both.
People with no creative talent might love them. It remains to be seen if these will basically turn into the next generation of (albeit dynamically generated) clip art.
Google also announced a partnership with Adobe Firefly to provide even more generative AI capabilities in generating images and other creations.
PaLM 2 Release
PaLM 2 is the next version of Google’s large language model.
The previous / current version, PaLM is the language model that powers Bard and other products. PaLM has four different-sized models (Gecko, Otter, Bison, and Unicorn) which are sized for various purposes and now supports over 100 languages.
It also boasts significant improvements in its ability to understand code, math, and a variety of writing styles.
Overall, Google seems focused on applications of AI that add value quickly, building on a lot of the paradigms popularized by the release of OpenAI products but leveraging their portfolio of assets to bring more value.
A lot of the agent and app frameworks we’re seeing in the open source world (including Langchain, BabyAGI, Auto-GPT, and LlamaIndex) are more generalized tools that represent the next evolution in generative AI: connecting LLMs to all of our data.
What Google is basically doing here is engineering products that do the same thing across their entire suite of products.
They have published their API and even if you don’t have access yet, you can get a head start on the types of things you can do.
Earlier this year, Google announced that Google’s Brain team and the DeepMind team settled their beef.
They are working together on a new model called Gemini.
The I/O keynote didn’t go into too many details but it sounds like this is the next-generation of model they plan to release after PaLM 2.
So, with every new announcement or technology release, SEOs are always asking themselves: “is SEO dead?”
Honestly, this industry sometimes sounds like a bunch of doomsday preppers constantly going on about the apocalypse.
As long as there is content and people searching for stuff on the internet, SEO isn’t dead (sorry, advertisers and VCs).
The relevant question, as always, is: how is the SERP going to evolve and how will that affect user behavior?
The biggest change is the introduction of conversation AI as, essentially, a SERP Feature.
What will be a big change is, what happens once that conversation starts?
As Google has demonstrated many times over, they have an interest in keeping searches on their properties for as long as possible vs. sending those users away to all of our websites.
So there will still be websites, there will still be links, but there is now a potentially much more engaging way to search.
As things evolve, the name of the game will be for site owners to ensure that their content and products are appearing not only in search results but also in the generated responses to content.
A lot of the traditional best practices for SEO will apply but there will likely be increased importance on structured data, knowledge graphs, and quality content.
Let us know what you think by chatting with us on Twitter or LinkedIn.