At a recent Google I/O - the search engine's annual developer conference - CEO Sundar Pichai announced a wide swath of AI-infused upgrades across all their web and mobile products. A major upgrade touted was the addition of AI Overviews to their search engine. Google was essentially integrating their generative AI model Gemini with their search engine to offer a 'snapshot summary' response to any given query. Ideally, it would save users the time and effort of having to sift through multiple sources themselves.
It was a fascinating development to Google's core search functionality, something that had not seen any major front-facing changes for years. However, the fanfare was short-lived: as it rolled out to US users, it became clear that the feature was not working as intended. Users quickly discovered that AI Overview was offering inaccurate information that was not just out-of-date or nonsensical, but could lead to fatal consequences. It was suggesting unsafe household practices, questionable medical advice, and more dangerous guidance. As one publication put it, 'Google makes 'technical improvements' to AI Overviews after they tell users to add glue to pizza, eat rocks, and drink urine.'
The New York Times noted that Google's fumble 'undermined trust in a search engine that more than two billion people turn to for authoritative information', and Google was forced into a mea culpa, claiming in a blog post that the errors occurred as a result of "data voids" - in other words, an information gap. Google acknowledged that, because of these gaps, AI Overviews might misinterpret the content online.
Google halted the rollout, but said it plans to reinstate the feature after ironing out the kinks and the early teething issues. For all the press, though, it is still not clear what safeguards Google will implement to prevent AI Overviews from going off the rails again.
This is already having a serious impact on the future of content marketing strategy.
First, current SEO approaches will no longer be fit for purpose, given that Google is shifting its own purpose. AI models can parse grammar and syntax reliably, but the finer nuances of language, such as tone and sarcasm, remain elusive and tantalizingly out of reach. Your audience has to look at other means of ranking and sorting content. More importantly, your content has the opportunity to break through by appealing in very human ways to your intended audience, but it can't just rely on search optimization. Luckily, AI can't replicate humanity yet.
Second, despite the short-term setback, AI is becoming a full-stack influencer from the content to the search engine. You will need to look at other means of assessing how audiences receive and interpret content, sifting through content which is becoming increasingly hard to differentiate between human-created and AI-generated. You need tools that take in SEO insights, but rely more on engagement, knowledge gaps, and more data-led information beyond search.
At Onside Content, we have created two proprietary indices around assessing your audience: the Likability Index, and the Consideration Index. Our Likability Index identifies where audiences are most engaged in topics related to your product or service, while our Consideration Index reveals where they are most curious. The former is indicative of current engagement, while the latter offers predictive potential to show what your audience is looking for next. Our novel two-pronged approach leans into the nuances and vagaries of language beyond what Google's AI Overview can currently offer.
We can upgrade your content marketing with such a considered approach. Onside Content is your bridge between data-led insights and the untapped audience you seek. Using our proprietary technology, we can help you find, create, and lead the smartest conversations in your space with a bespoke content strategy. We'd love to support! Drop us an email at info@onsidecontent.com.
