Surfaced - A Search Newsletter

The purpose of this newsletter is to surface insights that matter. Meticulously curated and not watered down, you’ll join me on my adventures through trying to make sense of the shifting sands on how AI is affecting digital marketing.

Expect thought leadership, POVs, data insights, and other tools and tricks to help turn you into an AI search champion. B2B, B2C, Enterprise, SMB - I’ll chew and digest for every

If I do this well, before you finish your coffee/tea you’ll be more informed and better prepared to handle the evolving search industry.

• Reddit this week continues to debate the measurement of AI search, showing how early we still are in the maturity. There are compounding gains for companies tackling this, though.

• A lot of data has come out recently about LinkedIn citations growing, which means more content will be fed to LinkedIn, which makes the platform continue to be noisy. 75% of my recent feed was determined to be written AI. Snake eating tail.

• Communicate to your content team that AI explicit will outperform implicit. LLMs don’t want to be Sherlock Holmes, they want to consume like executives.

LinkedIn is generating the most signal-dense content of any channel right now on AI search.

The big story this week: LinkedIn itself has become the #2 most-cited domain in AI search for professional queries, which is creating a never-ending feedback loop. Practitioners are posting about how to use LinkedIn for AI visibility, on LinkedIn, which then gets cited by AI.

Heads of SEO at B2B companies, VP-level marketing leaders, and agency owners are all posting their confusion. The Google March 2026 Core Update is also driving fresh urgency on LinkedIn for the SEO community, because they seemingly chopped down tons of AI aggregators that were causing spam (reported up to 80%). Yes, we will see the same efforts Google has been making for 25 years to combat low-quality answers in LLMs.

The Reddit signal this week is loud on one thing: measurement confusion is still the dominant pain for both SMB and Enterprise brands. Marketers who've spent years refining dashboards have no visibility into AI search performance, and the community is asking the same question in five different ways — "how do I know if any of this is working?"

The emergence of platform-specific citation fragmentation (ChatGPT vs. Perplexity vs. Google AI Mode, each pulling from different sources) is generating new frustration, with people’s heads spinning on how to actually measure them or (or if they actually need to look at all individually).

Data: Explaining the unexplainable to leadership

Everyone's losing their mind over zero-click search. Middle management is playing defense, and smart leaders are learning to report on what actually matters: revenue.

93% of sessions end without a website visit. I've seen the data at numerous companies. It looks really bad. But dig one level deeper, and the story flips from “the sky is falling” to “this is a hugely profitable channel”.

LLM-referred visitors convert at more than 2x the rate of traditional organic. Fewer people, but they're arriving pre-sold. AI already explained why they need you; they are basically an extension of your SDR team.

Here's the part no reader here wants to hear: 90% of what drives your brand's AI visibility is earned and owned media, largely it’s invisible to modern web analytics. We weren’t set up for this with traditional analytics, and measurement with AI visibility is more akin to brand marketing.

I’ll cover attribution and measurement in-depth in a later newsletter :)

Citation Gap

This is the gap between SEO and GEO in one chart. Traditional search rewards authority. Generative search rewards clarity, structure, and relevance. You can have both, but you have to build for both intentionally.

Here's the assumption most marketing teams are still making: if you rank well on Google, you're covered. The logic feels solid. LLMs are trained using Google search results. You're in the search results, so you're in the conversational answers. Right?

Nope. The data from Q1 2026 says otherwise, and the shift happened fast. In mid-2025, 76% of AI Overview citations came from Google's top-10 organic results. By Q1 2026, that number had flipped. 62% of citations now come from pages outside the top 10, and 80% of all LLM citations don't rank in Google's top 100 at all.

The pages AI engines trust most are, by traditional metrics, invisible. They're not winning on backlinks or domain authority or any of the signals your SEO tool is tracking. They're winning because they're structured for extraction, written to be cited, and optimized for how language models actually read content.

While Wikipedia and Reddit are cited more than the average domain, it’s not cause “domain authority”, it’s because it’s crawlable, valuable, and it covers so many topics.

The biggest takeaway: AI is a literalist. It prefers content that provides a clear 'map' of facts, as implicit content introduces a margin for error, which is why well-structured sites are favored.