Skip to main content
Back to Blog
ResearchFebruary 5, 2026

Introduction to Decision Boundaries in AI Search

Daniel Wang

Daniel Wang

Founder · UC Berkeley MIDS

TL;DR

AI models have a Decision Boundary — the threshold where they switch from ignoring your brand to recommending it. Sill uses Monte Carlo simulations to map these boundaries by testing thousands of prompt variations.

When you ask ChatGPT "What's the best project management tool?", it doesn't search the web in real-time. Instead, it synthesizes an answer from its training data and any grounding sources, making probabilistic decisions about which tools to mention.

This is the Decision Boundary—the threshold at which an AI model switches from ignoring your brand to citing it.

Understanding the Threshold

AI citation is driven by semantic authority, contextual relevance, and mention frequency — not traditional SEO metrics like Domain Authority or backlinks.

Decision Boundaries are determined by semantic authority, contextual relevance, and frequency of brand mentions — not traditional SEO metrics like Domain Authority.

Traditional SEO metrics like Domain Authority or backlink counts don't directly influence LLM recommendations. Instead, AI models evaluate:

  • **Semantic Authority**: How strongly your brand is associated with specific concepts
  • **Contextual Relevance**: Whether your offering matches the user's implied needs
  • **Frequency and Recency**: How often and recently your brand appears in relevant contexts

The Monte Carlo Approach

Sill runs thousands of prompt variations, adjusting one variable at a time, to pinpoint the exact threshold where AI citation probability shifts.

Sill maps Decision Boundaries by running thousands of prompt variations, adjusting one variable at a time to identify the exact threshold where citation probability changes.

At Sill, we use Monte Carlo simulations to map these boundaries. By running thousands of prompt variations and adjusting one variable at a time, we can identify the exact threshold where citation probability changes.

For example, we might find that a SaaS tool is cited 80% of the time when described as "enterprise-grade" but only 20% when positioned for "small teams."

Why This Matters

Knowing your Decision Boundary lets you optimize messaging, identify competitive semantic gaps, and measure AI visibility changes with precision.

Understanding your Decision Boundary lets you optimize brand positioning, identify competitive gaps, and track AI visibility changes with statistical precision.

Understanding your Decision Boundary allows you to:

  1. Optimize positioning: Adjust messaging to cross citation thresholds
  2. Identify gaps: Discover where competitors have stronger semantic authority
  3. Measure progress: Track visibility changes over time with precision

The future of brand visibility is probabilistic. The brands that understand this will win.

Daniel Wang

Founder · UC Berkeley MIDS

Previously at Nordstrom, Bloomberg, Hexagon (now Octave)