Skip to content

SEO Evolution & “Ranking” for LLMs

The SEO landscape is undergoing its most significant transformation since Google’s algorithm updates began reshaping search in the early 2000s. With the rise of Large Language Models (LLMs) and AI-powered search experiences like Google’s Search Generative Experience (SGE) and Bing’s AI-integrated results, the traditional playbook of keywords and backlinks is evolving rapidly.

The Shift from Traditional Search to AI Overviews

Search engines are increasingly presenting AI-generated overviews (AIO) and generative engine optimization (GEO) results that synthesize information from multiple sources before displaying traditional organic results. This means users often get their answers directly in the search results page, potentially reducing click-through rates to individual websites.

Instead of optimizing solely for position one on the SERP, we must now optimize for inclusion in AI-generated summaries and conversational search responses.

Strategies for LLM and AI Search Optimization

Focus on Entity-Based SEO

LLMs excel at understanding entities and their relationships. Structure your content around clear entities (people, places, concepts) and explicitly define relationships between them. Use schema markup to help AI systems understand your content’s context and meaning.

Embrace Natural Language Processing

Write in a conversational, natural tone that mirrors how people actually speak and ask questions. LLMs are trained on human language patterns, so content that reads naturally performs better in AI-generated responses.

Create Comprehensive, Authoritative Content

AI systems favor content that demonstrates expertise, authority, and trustworthiness (E-A-T). Develop in-depth resources that thoroughly cover topics rather than thin content targeting single keywords.

Optimize for Featured Snippets

While not new, featured snippets have become even more critical as they often serve as source material for AI overviews. Structure content with clear headings, bullet points, and concise answers to common questions.

Implement Structured Data

Rich snippets and structured data help LLMs understand and categorize your content. This increases the likelihood of being referenced in AI-generated responses.

New Metrics to Track

Traditional metrics like keyword rankings and click-through rates remain important, but new KPIs are emerging.

  • AI Visibility Score tracks how often your content appears in AI-generated overviews and summaries across different search engines.
  • Source Attribution Rate monitors when your site is cited as a source in AI responses, even without direct traffic.
  • Conversational Query Performance analyzes performance for long-tail, question-based queries that mirror natural speech patterns.
  • Entity Recognition measures how well search engines associate your brand with relevant entities and topics.
  • Zero-Click Impression Value which tracks brand exposure and authority building even when users don’t click through to your site.

Start Tracking LLM Referral Traffic

The first thing you can do is set up a report to track referral traffic that is coming from LLMs. This is pretty straightforward if you have a tool like GA4 already in place. Here is a step-by-step guide to creating that report within Google Analytics.

Step-by-Step Guide to Creating Your Report

Step 1: Navigate to the “Explore” Section

  1. Log in to your Google Analytics 4 property.
  2. From the left-hand navigation menu, click on Explore.

Step 2: Create a New Exploration

  1. In the “Explore” section, click on the Blank template or the “+” to create a new exploration.
  2. Give your exploration a descriptive name, such as “LLM Referral Traffic.”

Step 3: Add Dimensions and Metrics

To build your report, you’ll need to import the dimensions and metrics you want to analyze.

  1. Dimensions: In the “Variables” column on the left, click the “+” next to “Dimensions.” Search for and import the following:
    • Session source / medium
    • Page referrer
    • Landing page + query string
  2. Metrics: In the same “Variables” column, click the “+” next to “Metrics.” Search for and import the metrics you’re interested in. Good starters include:
    • Sessions
    • Engaged sessions
    • Engagement rate
    • Conversions (if you have key events set up)

Step 4: Create a Segment to Isolate LLM Traffic

This is the key step to filter your data to only show traffic from known LLMs.

  1. In the “Variables” column, click the “+” next to “Segments.”
  2. Choose Session segment.
  3. Name your segment something clear, like “LLM Traffic.”
  4. Click Add new condition.
  5. In the condition builder, select the dimension Session source.
  6. Change the match type from “contains” to matches regex.
  7. Paste the following regular expression into the value field. This expression includes many of the most common LLM and AI-powered search referrers (below)
  8. Click Apply.
  9. At the top right, click Save and Apply.
^.*(ai|openai|copilot|chatgpt|gemini|gpt|perplexity|bard|edgeservices|bnngpt).*$

Step 5: Build Your Report

Now you’ll use the dimensions, metrics, and the segment you created to build your report in the “Tab Settings” column.

  1. Segments: Drag your newly created “LLM Traffic” segment from the “Variables” column and drop it into the “Segment Comparisons” box in the “Tab Settings” column.
  2. Rows: Drag Session source / medium from “Dimensions” into the “Rows” box. You can also add Landing page + query string as a second dimension here to see which pages are getting the traffic.
  3. Values: Drag the metrics you want to see (e.g., Sessions, Engaged sessions) from “Metrics” into the “Values” box.

Step 6: Analyze Your Report

You should now see a table showing traffic data exclusively from the LLM sources defined in your regular expression. You can:

  • Identify Top Referrers: See which LLMs are sending you the most traffic.
  • Analyze Landing Pages: If you added the Landing page + query string dimension, you can see which specific pages on your site are being linked to.
  • Change Visualization: Use the icons at the top of the “Tab Settings” column to change the report from a table to a pie chart, line chart, or other visualization.

Next Steps and Considerations

  • Refine Your Regex: The world of LLMs is constantly changing. You may need to update your regular expression over time to include new and emerging AI platforms.
  • Create a Custom Channel Grouping: For a more integrated approach, you can create a custom channel group for LLM traffic in the “Admin” section of GA4. This will allow you to see LLM traffic alongside other channels like “Organic Search” and “Direct” in your standard acquisition reports.
  • Looker Studio: For more advanced visualizations and dashboards, you can connect your GA4 data to Looker Studio and build reports there using the same filtering logic.

The future of SEO lies in creating content that serves both human readers and AI systems. This means maintaining the fundamental principles of quality, relevance, and user value while adapting to new technical requirements and user behaviors.

The evolution isn’t just about ranking higher, it’s about becoming the authoritative source that AI systems trust and reference.

Tags: