A little over a year after Google first demonstrated Bard (now called Gemini), its answer to OpenAI’s ChatGPT, the company began integrating its generative AI technology into web search results, and you are probably aware of some of the fallout.

Last month, coinciding with Google’s I/O developer conference, the company rolled out ‘AI overviews’ to logged-in US users: generative AI summaries that answer the user’s search query without them needing to sift through search results and synthesise this information themselves (though links to sources are provided for anyone wanting to check the data or read further).

The SEO world has been speculating for months about how this launch could impact search traffic and search marketing strategy more widely. However, when Google’s AI overviews rolled out, the biggest story was the numerous silly, problematic and downright wrong answers that appeared: including advice to “eat at least one small rock per day” (query: how many rocks should I eat), or the now-infamous suggestion that non-toxic glue might help prevent the cheese sliding off your pizza.

Google was quick to put out a blog post, simply titled, “AI Overviews: About last week” in which it detailed the measures being taken to correct these issues, including limiting reliance on user-generated content such as Reddit posts and limiting the inclusion of satirical or humorous content.

Google’s response feels very knee-jerk, and the execution is clumsy.

SEO and ecommerce consultant Carl Hendy notes that although AI overviews (also called Search Generative Experience by Google) have been in testing for some time, “Google’s response [to the issues with overviews] feels very knee-jerk, and the execution is clumsy.

“The current AI overview implementation is slow and often returns low-quality websites that would not have featured in traditional search queries,” he adds. “These results often lack links, making it harder for users to visit websites, requiring additional prompts and more effort.

“While Google’s priority lies with their shareholders (which is understandable), they also have a responsibility to content creators. […] My hope is that Google will find a way to drive clicks to a broader set of quality websites while also keeping their shareholders happy.”

How search marketers can adapt to Google’s AI overviews

At least for now, AI overviews seem to be here to stay, so what should search marketers do, if anything, to adapt?

Optimise across organic and paid – but approach AI overview ads with caution

Google has wasted no time in unveiling an ad proposition for AI overviews, announcing at Google Marketing Live on 21st May that it will be testing search and shopping ads in AI overviews for US users, which will appear “when they’re relevant to both the query and the information in the AI Overview”. Existing Search, Performance Max, and Standard Shopping ads can all be featured in AI overviews.

“Despite initial accuracy concerns and deployment limitations to FAQ-style searches, the introduction of ads in Google’s AI overviews presents new opportunities for marketers,” says Beth Nunnington, VP Organic Media at Journey Further. “While ad volume may be constrained initially, these opportunities will likely expand as the technology matures and user adoption grows.

“Also, maintaining high-quality, relevant, and engaging content across organic and paid channels will undoubtedly play a crucial role in ranking for AI-powered search results.”

Is Reddit trying to have its cake and eat it with Google and generative AI?

Kat Sale, Co-Founder at House of Performance, sounds a note of caution with regard to ads in AI overviews: “With any tech we’ve seen before, it’ll be rubbish at first, then improve with time and usage. Definitely test it, but make sure everyone involved in the brand being advertised is ok with the possible outcomes.”

Don’t sweat the lack of metrics

Google has yet to make available dedicated analytics for AI overview results in Search Console, instead grouping them in with traditional search data. This lack of insight has frustrated many search marketers – given that it makes it impossible to tell if any traffic shakeups can be traced back to AI overviews, or whether any changes in tactics have moved the needle with overviews.

“Don’t sweat the lack of traffic insights; this will come in time, and no-one else has them so you’re not losing out to competition,” advises Sale.

“Search marketers need to be at the top of their game to ensure they’re competing on the right queries and not wasting ad spend on irrelevant stuff. As Google keywords get wider and wider matching, it’s imperative marketers analyse both search terms and themes across their campaigns to ensure they’re on the right searches and maximising any new trends.”

Focus on brand building and user needs

“Ensuring your brand is visible across web platforms through links from other websites, branded search demand, recommendations in discussion forums, and maintaining a positive sentiment will help you succeed in AI search results,” says Carl Hendy.

Additionally, providing “a clean UX, an easy-to-crawl website, and populating as much of Google’s Knowledge Graph as possible will make it easier for Google and other search engines to understand the relationships and topics related to your brand,” Hendy adds.

Nunnington says that although the lack of dedicated analytics poses a challenge for measurement, “marketers should continue prioritizing strategies that align with Google’s emphasis on quality, relevance, and user engagement.

“This includes producing well-researched, original content tailored to specific user needs, optimising for engagement metrics like click-through rates and time on page, and leveraging brand-building activities to enhance visibility and attract high-quality backlinks.”

Econsultancy runs ecommerce and digital marketing learning academies for global brands.