Adswerve Roundtable: Bard, ChatGPT and the Future of SERP

Subscribe to our monthly newsletter to get the latest updates in your inbox

Pat Grady, Brent Ramos, Lauren Stone & Jeff Stuart


On the heels of the much-deserved press that chatGPT has garnered over the last several weeks, Google's CEO recently dropped a post on The Keyword detailing how AI is coming to Google's search engine results page (SERP). Within the post, Sundar Pichai details how Google's latest AI technology, dubbed Bard, will be slowly integrated into its consumer products. 

This marks an all-new era where conversational AI will become increasingly present on the SERP and is slated to fundamentally change how we interact with search in new and different ways, but what are the impacts? Lucky for you, we’ve rounded up some of Adswerve’s best minds in search and data to share their thoughts and help analysts and marketers better forecast the potential shifts that an AI-augmented search experience could bring. 

The Future of SERP Real Estate

Jeff Stuart  Platform Engagement Manager – Search


Jeff Stuart

Platform Engagement Manager - Search

JS: Discussions around the SERP have historically been dominated by the theme of real estate: paid results vs. organic (remember when Google increased the number of paid results to four?!), or sponsored vs. organic product listings, site links, snippets, and more. Not anymore.

Now, the conversation takes a 90-degree turn, and begins to question the very nature of how paid and organic results will fit into a SERP when an AI-driven conversation sprawls out and above the fold. Many may be asking, what is the new paradigm? The answer to this question may come from looking closely at the graphics embedded in Pichai’s post (shown below).

AI driven conversation
image courtesy of Google blog

Examining the animation, we can see that the conversation dynamically pushes down the linked results as more text is added on mobile. Details on how it may integrate on Desktop are sparse, but if Bing Chat is any indication, we can safely speculate that the conversations with Bard may eventually involve some hyperlinked components, akin to a “further reading” type of source reference. If true, this might facilitate organic, and perhaps even paid results, to become integrated within the conversation itself. The eventual end result might look like a living, breathing (and sentient? kidding) SERP where the user’s query no longer returns a rich-but-static series of products and links, but a dynamic conversation with Bard that highlights these as the conversation itself further refines the original query.

Google’s flub during Bard’s unveiling earlier this month and the bizarre reports around Bing Chat’s AI behavior is indicative of the growing pains that all AI will continue to experience as the technology evolves, but shouldn’t diminish the potential and eventual changes that it will have in the world of search. Brands and agencies would benefit from taking a medium view on this technology and begin incorporating it into their strategic and tactical planning to not only capitalize on the moment when Bard and its peers become more democratized, but to also gain or maintain an advantage in this space compared to their rivals when the tech matures.

Furthering the True Nature of Search

Brent Ramos  Product Director – Search


Brent Ramos

Product Director - Search

BR: The nature of search is a return on requested information, and historically, we experience that return in a ranked index of links. While these links have progressively evolved with targeted accuracy over time, if we think about the process as a whole, that experience is still pretty archaic. 

With BARD or other AI coming into the SERP,  it offers a more human-esque experience that is both conversational and provides an accurate return of information (without the experience of sifting through links or 2-step click paths landing on website pages to sift even more). It’s a change that potentially offers a better experience, more accurate results, and a faster return of information. 

Looking at it holistically, the believability of a LaMDA (the language-model technology behind Bard) vs manual human point-and-click to website links may actually rate higher in the believability weight at scale. Even with the flubs and missed facts as it continuously learns along the way - we can assume it will only get better over time at a faster clip than our manual attempts, and even our own mental bias deciding what sites to sift and consume. 

But let’s shift from algorithm efficacy to the actual market, where we can focus from a venture and media lens. Here, Chatbot AI is nothing new. There are many firms out there with advanced capabilities and venture dollars, so the interconnectivity of making it usable at scale is where we will see the upcoming battles won or lost. 

SERP returns are one thing, but it’s important to keep in mind that they tangentially power applications for maps, video, local services, voice, eCommerce, biotech and more. It’s where the speed-to-market AI race should focus on delivering against. It’s these applications of AI that will dictate winners in the market, which require a ton of lift outside of algorithm efficacy alone, and into the dredges of consumerism. If you back down another layer behind that, you will see the tangled web of ad buying ecosystems, analytics, ad tech tools, and brands or agencies wrapped within the paradigm as well. 

If people are spending more time in the SERP getting information, website landing may take a back seat as consumers convert more within the channel directly or just grab information with a zero-click chatbot experience. We have seen this trend in social media, and the SERP is catching up. PPC models may see additional bid types emerge that include BARD formats, sponsored brand chats with buy buttons, maybe a pay-per-branded prompt model or there could be an entirely new channel created for AI advertising. All of this is search-nerd speculation, but we do know one thing: Technology will evolve; it rarely recedes back. Even if AI and things like the Metaverse do not pan out to what we currently think they will become, they will certainly impact the consumers' day-to-day in some form. We are on the precipice of a new frontier for search experience that should be grasped with excitement, yet always anchored to human usefulness, integrity, and responsibility which is where the immediate AI focus should shift.

Imagining the AI-Powered Ads to Come

Lauren Stone  Paid Search Consultant


Lauren Stone

Paid Search Consultant


LS: As we begin to think about the different text ad placements that an AI-Powered Chat feature can bring to Google and Bing, I can’t help but get excited about the different features and optimizations that will become available to those building out the ads.

When we think about a run-of-the-mill text ad, such as an RSA (Responsive Search Ad), we generally have full control over our ad copy, imagery, ad extensions, and more. As a marketer, I don’t think it's unreasonable for advertisers to expect AI-created ads to host some of those same core features that allow us control over what gets featured in our conventional text ads. With that said, search engines will have a responsibility to make sure that the ad copy is being utilized properly and in an accurate context. A baseline demeanor and informative style of communication will be critical for these ad placements and their success, especially given the emotionally-charged and very humanistic responses we have seen from the current chatbots.

Looking ahead, I wonder if we’ll see an experience similar to that of dynamic search ads. Instead of providing ad copy, what if we provide Bing or Google with a specific URL to crawl for our campaign? Then, this would be utilized when the chatbot presents an ad to searchers. I think the flexibility here could work better in the AI space, as the bot wouldn’t be married to the precise ad copy provided and could grab what was appropriate from the webpage. It isn’t unreasonable to think that these search engines could roll out multiple types of text ads, so advertisers could find what works best for them. 

Additionally, it’s exciting to imagine the type of data that will be available to us when optimizing our campaigns. As search term data is available to us in a run-of-the-mill text ad, imagine the value of a search space in which we could receive the bots’ full answer along with the search query that was inputted. This would also be an area to help advertisers ensure that their AI-ads fall in line with branding. Will it be possible for us to negate specific words or phrases from the chatbot when a chat ad is populated? These types of questions will need to be addressed.

While we are still waiting on the granular specifics from Google and Microsoft, what we do know is that as marketers we should be prepared to be flexible as the search world adapts to these technologies taking off. Regardless of how autonomous this new ad placement will be, marketers will still play a critical role, being the ones with the most knowledge of the brands we represent. With our historical knowledge of past campaigns and understanding of our client’s branding and tone, paid searchers will be ready to go to bat to make sure that the brands we represent are being properly portrayed and seeing positive results in this new ad space - something that an AI-powered resource won’t be able to replicate. 

Bet on Bard

Pat Grady  Solutions Engineer


Pat Grady

Solutions Engineer

PG: Full disclosure, I'm the resident curmudgeon on ChatGPT and, more specifically, on the tech-media hype around Google's "code-red" in response to OpenAI + Microsoft coming after their search dominance. 

If there is to be a winner in this race, my bet is on Google. I think the incumbent has an advantage, and Google has proven to be nimble and competitive when they need to be. They've also researched and experimented across many different digital channels and experiences. They've integrated critical business learnings into their products and processes. They're also simply a good bet, historically speaking. 

Google has been using language models to improve its canonical indexes and knowledge graphs preemptively, for over a decade. Rather than expose risky models to the public, instead, they curate the output of those models through much safer software development and product management lifecycles. The "code-red" response is, in my opinion, knee-jerk, and probably a bureaucratic blunder. Still, it could pay off in the long run and allows Google to accelerate innovation by upping its risk tolerance.

Both Google and Microsoft + OpenAI seem committed to transparency and ethics. Their desire to gain the AI advantage will fuel innovation on each side, but that competition is good. The media gets hyped, the public learns the nuance, risk tolerance goes up, innovation follows, and the public benefits from new and shiny things patented and sold by big tech. 

If you have any questions about how this plays into your long-term search strategy, please contact us. We’re happy to help!