TL;DR

Google engineer Lauren Clark confirmed in the "Search Off the Record" podcast that AI Overviews and AI Mode are intentionally isolated systems — separate from each other and from Google's main ranking pipeline. This explains why brand visibility scores often differ dramatically between these two AI surfaces. The practical takeaway: you need to track and optimize for each system independently.

Key Takeaways

AI Overviews and AI Mode run on separate infrastructure with different retrieval logic
A brand can have 80% visibility in AI Overviews but only 30% in AI Mode — and that's not a measurement error
AI Overviews leans heavily on traditional Google index rankings (top-10 results account for ~92% of citations)
AI Mode uses more Gemini-native sources, including direct crawl data and Knowledge Graph signals
Optimization strategy must fork: classical SEO for AI Overviews, entity authority for AI Mode

What Google Actually Said

In late April 2026, Google engineer Lauren Clark appeared on the "Search Off the Record" podcast and dropped a significant insight about how AI features work inside Google Search.

The core revelation: Google treats its AI models as black boxes. Because they can't fully predict or debug AI behavior at scale, they deploy these models only in named, compartmentalized use cases — AI Overviews and AI Mode each run on their own isolated infrastructure.

This isn't a temporary architecture decision. It's a deliberate design choice that allows Google to iterate, debug, and fix issues in each system independently. As Search Engine Roundtable reported, Google has intentionally decided not to let AI broadly influence core ranking signals.

Why This Matters for Brand Visibility

If you've been tracking your brand's AI visibility across Google's surfaces, you've probably noticed something confusing: the numbers don't match. Your brand might show up consistently in AI Overviews but barely appear in AI Mode responses — or vice versa.

This isn't a bug in your monitoring tool. Google just confirmed it's by design.

Three Practical Consequences

1. Measurement Must Be Per-Engine

Aggregating visibility scores across AI Overviews and AI Mode into a single number is misleading. These are two different retrieval systems with different citation logic. When your monitoring dashboard shows a single "Google AI visibility" score, you're averaging two fundamentally different signals.

The fix: track AI Overviews visibility and AI Mode visibility as separate metrics. Compare them over time independently. Most GEO monitoring tools already collect the data separately — the gap is usually in how teams interpret and report on it.

2. Optimization Strategy Must Fork

For AI Overviews, the path is familiar. AI Overviews pulls heavily from Google's traditional index. Research shows that top-10 ranking pages account for approximately 92% of AI Overview citations. That means classical SEO — strong rankings, structured data, comprehensive content — directly drives AI Overview visibility.

For AI Mode, the playbook shifts. AI Mode runs on a more Gemini-native pipeline with its own crawl sources and stronger reliance on Knowledge Graph signals. Optimization here means building entity authority: claiming and enriching your Knowledge Graph presence, structuring content in Q&A and comparison formats that Gemini handles well, and ensuring your brand entity is clearly defined across authoritative sources.

3. The Gap Is a Diagnostic Signal

Here's where it gets interesting for strategic analysis. If your brand has strong AI Overviews visibility but weak AI Mode presence, that's a specific diagnostic signal: your traditional rankings are solid, but your entity authority needs work. If the gap runs the other way — visible in AI Mode but not AI Overviews — your entity presence is strong but your page-level SEO needs attention.

This gap analysis becomes a practical framework for prioritizing where to invest optimization effort.

How to Run a Gap Diagnostic

A simple framework for using this insight:

Measure both surfaces separately for your top 10 brand queries
Calculate the gap: AI Overviews visibility minus AI Mode visibility
Interpret the direction:
- Positive gap (AIO > AI Mode) → invest in entity authority and Knowledge Graph
- Negative gap (AI Mode > AIO) → invest in traditional SEO and page rankings
- Small gap (±10%) → balanced presence, focus on expanding query coverage
Re-measure monthly — these systems evolve independently

Tools like BeRecommended can help you track per-engine visibility across Google's AI surfaces and other AI search engines, making this gap analysis straightforward.

The Bigger Picture

Google's isolated systems approach has a broader implication: there is no single "AI search" to optimize for, even within Google alone. Each AI surface is its own optimization target with its own citation logic.

As Google continues to experiment with AI in search — and as other engines like ChatGPT, Perplexity, and Claude build their own retrieval systems — the brands that win will be those tracking and optimizing visibility per engine, not chasing a single aggregate score.

The era of one-size-fits-all search optimization is over. The question isn't whether your brand is visible in AI search. It's which AI search systems see you — and which ones don't.

FAQ

Q: Does this mean traditional SEO is irrelevant for AI search? A: No — AI Overviews relies heavily on traditional rankings. But it's no longer sufficient on its own. You need entity authority work alongside classical SEO.

Q: Will Google eventually merge these systems? A: Based on Clark's comments, isolation is intentional and structural. Google wants to debug and iterate on each system independently. Convergence seems unlikely in the near term.

Q: How often should I check my per-engine visibility? A: Monthly is a good cadence for strategic tracking. Weekly if you're actively running optimization campaigns.

Google Confirmed: AI Overviews and AI Mode Are Isolated Systems — Why Your Brand Visibility Scores Differ