6 SEO Predictions For Google Bard

We’re only a short time away from the search landscape completely changing forever. AI chatbots are on the horizon as Google and Bing appear to be launching their products in the coming weeks.

Google CEO Sundar Pichai announced that a chatbot called Bard that will be released “in the coming weeks”. Bing has also announced plans to integrate ChatGPT into their search engine and some users have already spotted Bing testing the interface.

Related Content:

• B2B SEO Services
• Lawyer SEO Services
• SaaS SEO Agency
• Technical SEO Agency

So what does this mean for the search landscape as a whole? What impact will it have on SEO? While we can’t possibly know for sure, there are predictions we can make about how this will change the digital landscape. Below are some of my top predictions for what we can expect moving forward.

Prediction #1: The Chatbot Will Cite Sources Of Information

This insight was originally pointed out to me by the great Kevin Indig. Recently, I read an edition of this Growth Memo newsletter (a must-subscribe for marketers) that discussed “Early attempts at integrating AI in Search”.

In this article, he discusses how there are already early versions of search engines that directly utilize AI chatbots. He makes note of how chatbot technologies will likely need to cite their sources due to legal reasons and to provide comprehensive experiences.

This reminds me of how Google currently handles featured snippet results where they directly embed results in the SERPs but cite the article where they’re scraping the content from.

In Kevin’s article, he mentions several search engines that already are using AI chatbot functionalities and likens them to MVPs of what Google’s rollout could look like. One that he mentioned, You.com, caught my attention as a potential application of what Google’s chatbot could look like.

The report from CNBC mentioned how one of the potential designs of Google’s chatbot placed the functionality “directly under the main search bar, replacing the current “I’m feeling lucky” bar”.

This is similar to how it’s structured on You.com’s home page:

If we consider You.com to be an early AI-chatbot MVP, then it can at least serve as a guiding light of what this functionality might look like in platforms like Google. As a result, I decided to perform different tests using the platform to see what insights could be gained.

Here’s an example result when I type “What are the best running shoes”. As Kevin noted, you can see how facts within the results are cited both in the text and in a references section below the answer.

To further provide evidence of this, Google’s initial demonstrations of Bard show a “Read more” section in mobile devices. Here it appears to pull the featured image, logo, title tag, and a description directly into the answer:

This is extremely important because this might lay the foundation for what “visibility in a chatbot” looks like. Two basic areas of opportunity would be to be included in:

1. The answer itself
2. The citations at the bottom

If your brand is wondering how to be visible in a chatbot platform, this might be the initial method (although I suspect there will be more). Having your site referenced in the answers might be the primary way that a chatbot drives both brand awareness and site traffic.

Prediction #2: Chatbot Answers Will Be Heavily Influenced By Aggregator Sites

So the next logical question becomes, “How could my website be the one Google chooses?”. Let’s go back to the previous example of a question for “What are the best running shoes?”

The chatbot tells us that the best running shoes are “Nike Air Zoom Pegasus 39”, “New Balance Fresh Foam X 1080” and “Brooks Glycerin 20” and cites the results directly in the answer itself via annotations and in a bibliography-style section at the bottom.

When I navigate to the citations it’s referencing, I’m pointed to an affiliate article on Tom’s Guide that references these exact three models as the best shoes.

Unsurprisingly, this article ranks on the first page of You.com (albeit at the bottom) for “best running shoes”.

In another example, I asked You.com what the best BBQ restaurant in Raleigh is. It tells me “The Pit Authentic Barbecue”, “Ole Time Barbecue”, and Clyde Cooper’s Barbecue and references articles from TripAdvisor and Yelp.

Going into these sites like TripAdvisor, it appears that the #1 and #2 businesses You.com recommends are the exact same.

From the tests I’ve run, it seems that You.com is likely to pull answers from top-ranking content it finds on aggregator sites on the first page. Clearly, it needs some data source to pull the information, so articles it deems as trustworthy enough to show users seems like a logical conclusion.

If this is the case with a search engine like Google, it will give aggregator and affiliate sites even more power than they have today. There’s already been a huge shift in search engines to favor aggregator sites and a chatbot that is routinely required to stitch together information might amplify that. This is especially concerning in the cases of affiliate sites where the products reflect those where creators can earn commissions and not necessarily the best products on the market.

However, this could be a new reality of search. Sites that have partnerships with affiliates that tend to have strong “barnacle SEO” on third-party platforms (think Yelp, G2, Trip Advisor) will have a competitive advantage because search engines lean on these sources to deliver answers.

However, I believe Google will be able to go “deeper in the results” and be more likely to serve information “past the first page”. Remember passage indexing? Serving snippet answers might be the perfect use case for this implementation.

Prediction #3: Title Tags & Images Will Be Important For Chatbot Click-Through Rates

If a Google chatbot cites sources, it will need to pull some type of page-level information directly into the interface. Currently, on You.com, each citation is pulled in via both the title tag and URL. Here’s an example of when I asked it “macbook air vs pro”

While there’s no telling exactly how Google would pull this information, given their longstanding preference of the title tag, it’s reasonable to assume that it will come into play. This means that “clickable” title tags and URL structures could become more important than ever.

Your site may no longer be competing for clicks with 10 other sites and a bunch of rich results. Instead, you could be competing against a much more limited set of results. This means that the result with the title and URL best optimized for the question would be the winner.

In the example above, notice how Apple is not well positioned in the result. While CNN and Forbes have title tags that directly relate to the question, the Apple result does not talk about the differences in the title tag and instead uses “Mac – Compare Models | Apple” as the title tag and a parameterized URL. As a result, they might miss this opportunity for traffic/conversion despite being included in the answer.

As well, images will play a crucial role in click-through rates from chatbot answers. Going back to Google’s example, notice how the first result contains a featured image. Compare this against the Quora result that is only text-based.

We’ll want to watch closely as to how Google presents these citations. If it scrapes title tags and URLs directly into the chatbot, marketers will need to consider these elements from both ranking in the traditional results and a CTR perspective in the chatbot.

Prediction #4: Search Behavior Will Shift Towards More Long-Tail Keywords

One of the most interesting changes a chatbot application would bring is how it impacts consumer search behavior. How many people will opt to interact with the chatbot instead of the traditional results? Will users use both and if so, how does their search behavior change when they leave the chatbot and enter the results?

ChatGPT-like bots seem great at giving users informational answers to their questions. However, they lack the ability of a traditional search engine to perform in-depth research.

For instance, if researching a product, a user might use the chatbot to gain a general understanding and then shift to the search engine to perform deeper product research. There they could find content a chatbot isn’t capable of producing as quickly such as reviews, price comparisons, product details and much more.

For example, the user journey could look something like this:

1. Chatbot question #1: “What are the types of running shoes?”
2. Chatbot question #2: “What are stability shoes?”
3. Search: “best stability shoes”

This is an overly simplistic example but hopefully it illustrates how consumers might interact with hybrid search. Using the chatbot to gain an informational understanding and then going deeper with more specific longer-tail queries that a chatbot cannot efficiently answer for them.

One way that marketers can look at this will be using Search Console data. If a chatbot is to rollout in the near future, an interesting exercise will be to compare query data from before and after the date of implementation. What types of queries are seeing significant drops in clicks and impressions? Are there queries that users tend to search more often now?

There will be a shift in how people use traditional search engines and Search Console will be an invaluable data source for marketers everywhere.

Prediction #5: Google Will Plan To Make Results More Visual, But Serving Costs Will Slow It Down

While our interactions with ChatGPT lead us to see a chatbot query as primarily text-based, I predict that Google will plan on changing that. Over the years, we’ve seen Google make aggressive pushes to search to give users a much more visual experience. Popular products, playable videos, new top stories layouts the list goes on and on.

This leads me to believe that it’s not a matter of if but when Google makes chatbot results more visually engaging for users. This means that elements such as image quality, reviews, and structured data would become important from a chatbot visibility perspective.

The only limiting factor would be the expense of incorporating these different elements into their platform. In Kevin’s article, he mentions how “ChatGPT costs $100,000 a day to maintain for one million users”. While that’s pennies to the search giant, imagine what those costs will be when you significantly amplify the number of users and queries that are processed.

A key piece of information in the Bard announcement was that Google will be releasing a “lightweight model version of LaMDA. This much smaller model requires significantly less computing power, enabling us to scale to more users, allowing for more feedback”. This is perhaps the biggest limitation to Google rolling out the chatbot and continuing to iterate new features. It will come at a significant expense to run the platform and integrate other forms of Google search.

Prediction #6: Brand Will Become Even More Important Than Ever

As we mentioned earlier, aggregator sites may play a huge role in the answers created by an AI-driven chatbot. However, Google has a huge advantage over less robust search engines like You.com. Google’s Knowledge Graph represents its inherent understanding of information that’s separate from the contents of an individual page. It serves as Google’s “internal encyclopedia” of the data it’s collected.

A simple example of this is a search for “christmas movies”. Using data from the Knowledge Graph, Google serves movie thumbnails directly in the search results:

Here’s another example when I search for “marketing automation platforms”. In the “Related Searches” section at the bottom, Google pulls in the logos of technologies that fit in this topical category. This is likely pulled from their Knowledge Graph:

The Knowledge Graph is something that can’t be influenced as easily as the contents of an individual page. It’s a collection of information pulled data sources a variety of data sources such as:

1. Content on your website
2. Third-party references
3. External links
4. Consumer search behavior
5. Trust signals
6. Review data
7. Wikipedia data

Overall, it’s a reflection of how strong an individual brand’s digital presence is.

If someone asks Google’s chatbot “What is the best marketing automation platform?”, my prediction is that Google’s Knowledge Graph will help inform that answer. The brands that have been investing in their digital presence for years and associate their website with that topic will be the winners.

This is why it’s important when Google talks about having site expertise. If Google truly views your site as the authority in that particular area it’s going to be more confident in showing your brand directly to users.

When it comes to the crucial moment of users asking a chatbot directly what the best products are, your brand strength will be a crucial component of Google’s answer.

Conclusion

While we can’t possibly know what the exact future holds, we can make educated predictions about what the future of our industry looks like with these new means of information delivery. It’s an extremely exciting time in our industry and perhaps a bit scary too. However, the brands that embrace the change of new technologies and search behaviors will put themselves in a great position moving forward to come out ahead of the pack.

About Chris Long

Chris Long is the VP of Marketing at Go Fish Digital. With 10+ years of SEO experience, Chris works unique problems and advanced search situations to help clients improve organic traffic through a deep understanding of Google's algorithm and web technology. Chris has advised on the search strategies for clients such as GEICO, Adobe & The New York Times. He is a contributor for Moz, Search Engine Land, and Search Engine Journal. He is also a speaker at industry conferences such as MozCon, SMX East, BrightonSEO and more.