We’ve been talking about the impact of AI on organic search, but it seems like in the last week I’ve had more conversations about it than in the entire year prior.
Perhaps we as an industry have suddenly caught up to the idea that we’re reaching a tipping point.
Everything is changing so fast, and it’s easy to feel overwhelmed. I’m here to help move the conversation from where things are to what needs to be done in response.
Last week, two articles made the rounds in our offices. The Wall Street Journal, using Mailchimp research, predicted websites’ primary function would be AI data sources. Semafor wrote on the importance of journalism in training AI bots.
They raise a few questions: Will anyone come to my site? How can I prove that my marketing is doing anything if they don’t? Do we need to keep investing in SEO? Is my brand showing up in ChatGPT? How can I know and give it the best chance of appearing there?
Has AI overtaken organic search?
The data shows a clear shift.
- Two thirds of Google searches stay in Google’s ecosystem – only one third go out to another website.
- Google searches fell in Safari for the first time ever in April.
- Desktop outbound CTR can fall by two-thirds the moment an AIO appears.
If organic search traffic to your website hasn’t already started falling, it will soon.
Are financial services an exception?
Yes, AI Overviews are becoming much more prevalent for searches. But that doesn’t mean all searchers naturally trust them for every search. Kevin Indig’s UX study of how AI Overviews impact search behavior found that searches with high-risk outcomes (what we call your money or your life searches) that typically include financial-related searches cause audiences to dig more into the search for validation. That means that for those of us working in the financial services industry, when the stakes are often higher, audiences are less likely to trust an AI answer off the bat and more likely to click through to a website.
Google Search Console data for most of my clients shows the very beginnings of these trends. We can see a very slight indication of decreasing clicks and more impressions in the last three months. That validates both that AI Overviews and search volume overall are increasing and that the higher bar for trust in the financial industry means a shallower decline in web traffic.
That might give us a reason to maintain a strong web presence, but either way, we can’t ignore the fact that AI-enabled search is here to stay, and it’s likely only going to become more influential. And that means we have to find ways to influence it.
How to get your brand into the AI results
There’s two common themes in all the research and opinions I’ve seen about how to influence AI results:
- Trust and reputation
- Technical signposting and semantic structure
The Semafor article found earned media in authoritative publications was the primary driver of how GenAI platforms recommend and describe brands and products. Avenue Z found that sites like Nerdwallet, Business Insider and CNBC were cited as sources in almost half of AI searches relating to fintech and payments topics inside ChatGPT, Perplexity and Claude.
To be visible in AI outputs, companies need to influence the language used to talk about them online, beyond just their website. This means traditional PR efforts like podcasts, YouTube, Reddit, third-party reviews, and earned media coverage are now core SEO assets.
It’s not enough to do really great PR. There’s a big push in the SEO community towards SEOs mastering a deep technical knowledge of things like RAG, vector embeddings and relevance engineering. We used to talk about how to optimize websites and pages. Now, pages are chunked semantically and indexed so that LLMs can retrieve and synthesize passages in response to queries, so we need to focus on how we optimize sections and paragraphs.
What that ‘technical signposting’ actually looks like isn’t all that different from the technical best practices we’ve talked about for years. It means structuring content with clear semantic HTML, implementing FAQ schema markup for common questions, and organizing content into discrete, self-contained modules that answer specific questions completely. It’s also applying these best practices to expert content and the about page and team bios, as these form your ‘public knowledge graph’ and are the training data for LLMs.
Both of these so-called ‘AI ranking factors’ mean that the companies that have been doing it well for years are likely to be the ones who continue to succeed. Brands with strong recognition, big PR budgets and great websites – like Blackrock, JP Morgan and PayPal – will likely have an advantage over lesser known fintechs and smaller investment managers.
What should you do about it
Like we said, nobody knows exactly how the AI models work and they are constantly changing. What we do know is our best guess based on observable trends, and that gives us at least a place to start.
Start with auditing and monitoring your AI visibility. There’s an abundance of tools popping up that claim to be able to do this. At its most basic, you can ask your AI model of choice to list 10 entities it associates with your brand. Prioritize earned media and make sure that your PR team and SEO/content/marketing teams are talking to each other.
Marketing has always been an influence game. We’ve always been trying to find where our audiences spend time crafting the message that resonates. The places and sources of influence might be changing – and changing fast – but the fundamentals of why we do what we do aren’t any different. If we can adapt quickly, I’m confident we’ll find new ways to have a voice, but if we cling to outdated SEO practices we’ll find ourselves increasingly invisible.
Hannah Porritt is a vice president in New York
