Why AI Can't Find You
(and How to Fix It)
The
Invisible
Audience
MARIA DYKSTRA
Last reviewed: June 26, 2026 · Forthcoming October 2026
Coming October 2026

The Invisible Audience

Why AI can understand your category and still skip your company.

Your first audience is no longer human. Before a buyer encounters your work, AI systems classify, summarize, and recommend it. Your homepage. Your founder bio. Your LinkedIn profile. Your reviews. From those fragments, AI decides who you are, who you serve, and whether you belong in the answer.

The Invisible Audience introduces the Algorithmic Authority Stack, a seven-layer diagnostic for B2B AI visibility. Built from 70 entities diagnosed across 50 client audits and the 20-company Algorithmic Authority Index. Specific failure patterns. Real data. Direct diagnostics you can run today.

From Chapter 5 · The Identity Resolution Failure
B2B fintech VP · six surfaces, six descriptions
Website: "platform"
LinkedIn: "solutions provider"
Crunchbase: "financial technology company"
G2 profile: "payments partner"
Sales deck: "the operating system for modern finance"
CEO bio: "fintech infrastructure"
How AI filed her
Ambiguous
Her competitor · half her revenue, a quarter of her team
Used "B2B payments infrastructure" on every surface. Every time.
How AI filed the competitor
Category leader

Seven structural failures. One stack.

Across 70 B2B entities diagnosed, the same patterns repeat. The Algorithmic Authority Stack names where the signal breaks. Each layer is independently testable. Each one has a different fix.

Layer 1 · Market Identity Clarity

Identity Fragmentation

Your homepage, LinkedIn, and About page describe you differently. AI cannot resolve which version is the real company.

AI does not see range. It sees contradiction.

Layer 2 · Expertise Architecture

Authority Collapse

Twenty years of real expertise. No structured proof AI can verify. The claim becomes noise. The work disappears into the category.

AI reads your claims without enough reason to repeat them.

Layer 3 · Semantic Density

Semantic Drift

Fifteen terms for one concept across your surfaces. AI splits your authority across every variation. None of them gather enough weight to land.

Your authority does not compound. AI cannot connect the evidence.

Layer 4 · Training-Ready Content

Citation Invisibility

200 posts. Three citation-worthy passages. AI evaluates content paragraph by paragraph for extractability. Narrative depth without structure produces invisible content.

Publishing more into a broken architecture creates more invisible content.

Layer 5 · Algorithmic Touchpoint Presence

Surface Dependency

You dominate ChatGPT and are entirely absent from Perplexity. Each platform reads different surfaces. Only 11 percent of cited domains overlap across the major AI systems.

Platform-local visibility is not AI visibility.

Layer 6 · Trust and Proof Signals

The Trust Gap

Your own pages talk about you. No independent voice does. AI pattern-matches entities across the web before citing them. A perfect page cannot compensate for a broken entity.

Google evaluates pages. AI evaluates entities.

Layer 7 · Visibility Measurement

Measurement Blindness

Your dashboard is green. Your Share of Model is zero. Both can be true at the same time. The dashboards your team trusts cannot see the metric that decides whether AI recommends you.

If you cannot see the citation gap, you cannot close it.

What you are likely missing.

Each insight below is the kind of finding that lands wrong inside a B2B marketing team. They sit on dashboards no one is looking at. Each one is the subject of a specific chapter in the book.

That 67 percent of pages cited in Google's own AI Overviews do not rank in Google's top 10. Ranking and citation are different contests now.

Chapter 2Ranking Is Not Selection

That AI crawled your site nearly 71,000 times for every visitor it sent back. Your insight feeds the answer. Your name does not appear in it.

Chapter 3The Crawl-to-Cite Problem

That AI files your company as "ambiguous" when your homepage, LinkedIn, and About page describe you differently. Your competitor with one consistent description was filed as the category leader.

Chapter 5Market Identity Clarity

That adding specific statistics and named sources to existing content lifts AI citation rates by 30 to 115 percent. Princeton measured it. Most teams are publishing assertion without evidence.

Chapter 6Expertise Architecture
Chapter 8Training-Ready Content

That AI splits your authority across every synonym you use for the same concept. One anchor term used across every surface beats ten variations every time.

Chapter 7Semantic Density

That you may dominate ChatGPT and be entirely absent from Perplexity. Your buyers use both. Only 11 percent of cited domains overlap across the major AI platforms.

Chapter 9Algorithmic Touchpoint Presence

That AI evaluates entities, not pages. A perfect page-level E-E-A-T score does not produce AI citations if the broader web does not corroborate the same expertise on independent surfaces.

Chapter 10Trust and Proof Signals

That your dashboard is green while your Share of Model is zero. 72 percent of AI citations produce no referral click. You are measuring the last 20 percent of the path to purchase.

Chapter 11Visibility Measurement

The book

14 chapters. Three parts. The diagnostic, the framework, and the sequenced fix. Built from 70 B2B entities and 420 platform-level assessments. Every chapter ends with a diagnostic you can run on your own company in under twenty minutes.

Part I · The Problem
Chapter 1

Why Companies Become Invisible

The competitor in your category that AI cannot find

Chapter 2

Ranking Is Not Selection

Why number one on Google can lose the AI citation contest

Chapter 3

The Crawl-to-Cite Problem

AI reads everything you publish and sends almost nothing back

Part II · The Framework
Chapter 4

The Algorithmic Authority Stack

The 7-layer diagnostic for AI visibility failure

Chapter 5

Market Identity Clarity (Layer 1)

The company AI cannot name

Chapter 6

Expertise Architecture (Layer 2)

Why your expertise does not add up to AI

Chapter 7

Semantic Density (Layer 3)

When every page describes you differently

Chapter 8

Training-Ready Content (Layer 4)

Why publishing more can make you harder to find

Chapter 9

Algorithmic Touchpoint Presence (Layer 5)

The platforms AI looks at before it cites you

Chapter 10

Trust and Proof Signals (Layer 6)

The proof AI can actually use

Chapter 11

Visibility Measurement (Layer 7)

Why your dashboard cannot see your invisibility

Part III · The Diagnostic
Chapter 12

The Algorithmic Authority Audit

How to read your own diagnostic

Chapter 13

What the Data Shows

70 entities. 5 industries. 420 platform-level assessments.

Chapter 14

The 30-Day Fix

Which layer first. Which engagement to run yourself. Which to hand off.

About the Author

Maria Dykstra is an AI Visibility Architect. She diagnoses why B2B companies are invisible to AI systems. She spent 13 years at Microsoft on the platforms that decided what reached buyers at scale, including a final role on product roadmap for media selling tools driving $2 billion in annual ad revenue across 36 global markets. She founded TreDigital in 2012 and ran it for 13 years across Fortune 500 and growth-stage B2B firms.

She created the Algorithmic Authority Stack after diagnosing 70 B2B entities, 50 in client audits and 20 in the Algorithmic Authority Index, a systematic study of AI visibility performance across ChatGPT, Perplexity, and Gemini. Her research, the Algorithmic Authority Index, examines how AI systems classify, retrieve, cite, and frame B2B companies.

Full bio →

What changed in AI retrieval this month.

One brief. The patterns your competitors aren't tracking yet. Covers ChatGPT, Perplexity, and Google AI Overviews. Published monthly.

Monthly. No spam. Unsubscribe anytime.

Do not want to wait for October?

The book teaches the diagnostic. The Snapshot runs it on your company.

The AI Visibility Snapshot tests your company against the same 7-layer Algorithmic Authority Stack used in the book. Written diagnostic delivered in 48 hours. The Stack layer breaking first, named. Top competitors holding the slots you should hold, identified.

Run the Snapshot · $497 →