Hi there,

You need to know that…

I'm pausing my newsletter for the next few months to focus on a big project at NSI.

Thank you for taking the time to regularly read my thoughts and opinions on SEO, content, and marketing. I appreciate your support and I look forward to seeing you again in the future.

Enjoy today's stories.

Nicole.

P.S. You can contact me any time by replying to one of my newsletters, messaging me on LinkedIn, or on my website.

I want to share your:

Big news, latest resource, or outstanding results.
I'd love to support you by featuring you here. It's FREE!
​
Smash the button and show me what you've got.

Insights

Do This Before You Publish More Content
​

​
One statistic from a survey of US-based marketing and PR professionals jumped out at me this week.

42% of marketing teams responding to AI search are doubling down on scaled content production, often with the help of AI content generation tools.

Only 23% say their primary approach is refining and optimizing existing content to improve how AI systems interpret and cite it.

I understand the temptation because publishing feels like progress, and AI makes it easier than ever to produce more pages.

The problem is we know more content is not always better content. And, we've all seen that scaling content can deliver a short-term lift before eventually dragging down the performance of an entire site.

The first question for brands and teams looking to increase AI visibility is this:

What content on your website deserves another look?

Most brands have years of content sitting on their site and scattered across the web.

Old blog posts, outdated product pages, press mentions that no longer reflect reality, and profiles on third-party sites that say something slightly different from what your homepage says now.

All of that shapes how people and AI systems understand your brand.

AI systems use these pages to build a profile of your business, and if that profile is confusing or out of date, creating another 20 articles will not fix the problem.

Audit what's already there and make the best content better.

It creates opportunities for Digital PR, earns more brand mentions and gives journalists and industry publications a reason to cite your work instead of everyone else's.

Don't write anything new until you review what's already out there, whether it still holds up, and whether it tells one consistent story.

When was the last time you looked at everything your brand has said about itself over the years?

​
​The survey results: You'll either nod in agreement or be shocked by some of the results in this AI search survey.

Related: Using SEO to Correct How LLMs Describe Your Brand​

AI search is not static

A month ago, I discussed a SISTRIX analysis that found when ChatGPT switched to GPT-5.5, nearly HALF of all citations were redistributed in just 48 hours.

This happens even when nothing about your content changes.

Just last week, David Konitzny (GEO Researcher at Peec AI) published research to back up SISTRIX's findings on how LLMs silently alter how they search for and prioritize information.

The conclusion is that these internal model updates impact brand visibility in AI search results.

3 notable shifts in ChatGPT's retrieval behavior

• The "official" bias: ChatGPT is increasingly prioritizing authoritative and official sources by adding terms like "official site" or "official docs" to its internal queries.
  ​
• arXiv vs. Wikipedia: arXiv.org has seen a 300% increase in its share of retrieved domains since April 2026, while Wikipedia’s share plummeted from 40% to 7% in the same period. arXiv often acts as a "silent source."
  ​
• Precision retrieval in ChatGPT 5.6 Luna: The use of the site: operator in Query Fan-outs surged from 0.004% in version 5.5 to 43% in Luna. The model now uses initial searches to identify a domain and subsequent loops to find specific, quoted data points within that domain.

Implications for your content strategy

AI models appear to be moving from broad discovery to precision retrieval.

Here are the best tips to stay visible:

• Focus on first-party content: Models use iterative loops to target original sources, so having authoritative, structured data on your own domain is critical.
  ​
• Structured factual information: Content optimized around exact attributes (e.g., pricing, technical specifications, or unique service details) is more likely to be retrieved by the model's new Boolean-style search patterns.
  ​
• Continuous monitoring: AI search is not static which means brands must look beyond the final output and track the internal retrieval behaviors (like query fan-outs) to understand why their visibility is changing.

Original data as a GEO moat

Similarweb's latest article discussing original data as a GEO moat gives us a quick checklist for packaging original data so it earns citations:

• State the number, not the theme. Lead with the figure itself, not a sentence describing the fact that a figure exists.
  ​
• Attach a source, sample, and date to every number. “2.5x more likely to visit, US Desktop, Jul-Dec 2025” is citable. “Data shows a strong lift” is not.
  ​
• Put your best stat in a standalone sentence or table row. Don’t make a model dig through prose to find it.
  ​
• Publish the same finding in more than one place. A full report, a summary blog post, a social post, and a PR push to trade press, all pointing to the same number, multiply the domains that can cite it.
  ​
• Give it a name people repeat. “The Downstream Impact of AI Visibility” is easier to cite by name than “our latest marketing research.”

Do you follow me on LinkedIn? I share regular tips and stories I don't have room for here. Come and join me​.

Information

✶ Searching for answers

3 highlights from the latest YouGov report on how people use search engines, AI assistants, and other platforms to find information online.

1. Low trust levels

Only 28% of U.S. online searchers say they trust information from an AI assistant. This compares to 70% of respondents who trust a search engine.

2. Search engine dominance​
​
Traditional search engines remain the primary gateway to the internet, used by 86% of online searchers in the past 30 days.

They are the default starting point for nearly every information task, from product research (62%) to asking specific questions (69%).

3. The "Frequent AI Searcher" profile

A distinct group of high-frequency users has emerged, representing 15% of American searchers.

This group is predominantly:

• Male (56%) and Millennials (36%)
• Full-time workers (75%)
• Highly likely to verify information across multiple sources (44%)

What I learned

Want more?

Watch the YouGov recording: The new search journey: How AI is changing online.

Download the report: Searching for answers: How AI is changing online discovery in ​2026.

✶ Do you need to track your brand or name on the web?

Rand Fishkin gives you 7 excellent reasons to track your brand mentions or personal name across the web. He also shows you how to do it.

His video overview is worth 5 minutes of your time.

✶ Google teases Gemini 4

A sentence buried in Google's introduction to their new 'Flash' series of models stated:

If Google follows its normal 6-month cycle, the new model could launch before the end of this year.

Inspiration

Your Superpowers

Superpowers you can choose:

• Ability to change yourself & your mind
• Not taking things personally
• Not needing to prove you’re right
• Careful selection of all relationships
• Staying calm
• Being alone without being lonely
• Being ok with being uncomfortable
• Thinking for oneself

Credit to fs blog: Brain Food​
​

That's all for now. In case you missed my message above:

I'm pausing my newsletter for the next few months to focus on a big project at NSI.

Thank you for taking the time to regularly read my thoughts and opinions on SEO, content, and marketing. I appreciate your support and I look forward to seeing you again in the future.
​

In the meantime, feel free to ask me a question, send an interesting link, or tell me what's on your mind. I read all your emails!

Keep in touch