GWG › Research › T-Model
Framework v2.0 · Data Verified

T-MODEL

AI Cognitive Stratification Framework
Five stages of AI adoption — from T0 to T4

Last updated: June 19, 2026 · By Wilson
27%
20%
22%
15%
4%
T0T1T2T3T4

US adult population ~260M · Quinnipiac / Menlo / Anthropic / Pew 2025-2026

Scroll

Why the T-Model?

AI users are not a monolith. One strategy cannot cover everyone.

27%
Never used AI
61%
Used AI (6 mo)
19%
Use daily
3%
Pay for AI
The problem with traditional segmentation: Age, income, and occupation are no longer enough — a 25-year-old developer could be T4 while a marketer the same age is T1-T2.

What the T-Model solves: Segment by AI cognition depth and behavior → "Who is where → What they need → How to reach them."

Five Layers at a Glance

Each tier is a distinct market with its own product logic

T0 · Outsider 27% · ~70M people
Never used AI. Declining 6pp/year but slowing. Distrust > disinterest. Don't mention AI — sell the outcome.
T1 · Triers 20% · ~52M people
Tried once, didn't stick. No aha moment. Most fragile segment — could flow into T2 or back to T0.
T2 · Free Users 22% · ~57M people
Monthly active, free tier. Don't lack willingness — lack a "worth paying for" scenario. Largest convertible pool.
T3 · Power Users 15% · ~39M people
Daily/weekly active, "ask AI first" muscle memory. The paying base. Pricing anchor: save 1 hour = worth the monthly fee.
T4 · Builders 4% · ~10M people
Multi-tool + API + Agent. Highest ARPU, shape industry discourse. 4% influence the other 96%.

Sources: Quinnipiac 2026.3 · Menlo Ventures 2025.4 · Anthropic 2025.9

Six Personas

Name, age, occupation, behavior — a real person in every tier

M
Marcus
31 · Software Engineer · Builder
Multi-tool stack, API + Agent, needs control
T4
P
Priya
29 · PM / Freelancer · Efficiency Payer
Daily user, subscribed, wants less switching
T3
J
Jordan
34 · Marketing Coordinator · Convertible Casual
Monthly free user, needs a high-frequency hook — biggest growth pool
T2→T3
S
Sofia
20 · Student · GenZ Native
Learning + social + companionship, free tier, high reach
T1-T3
D
David
52 · Mid-Manager · Waiting Professional
Heard of it, hasn't tried. Needs embedded tools
T0-T1
A
Andrea
58 · Clerk · Outsider
Not a target market yet. Do not invest now.
T0

Four Cross Dimensions

Look beyond tiers — who uses it, what for, and how big the gap is

Use Case
Ask (info/advice) 49% · Do (produce) 40% · Express (share) 11%
OpenAI / NBER 2025.9
Occupation Gap
White vs blue collar 2.5-3x
Software/IT: 36% of Claude usage
Anthropic 2025.9 / Gallup
Age Divide
18-29 58% · 30-49 41%
50-64 25% · 65+ 16%
Pew Research 2025.6
Topic Distribution
How-to 29% · Research 24% · Writing 24% · Code 4%
OpenAI 2025.9
Gender Gap
M > F by 8-13pp, narrowing
Men: coding, Women: writing
St. Louis Fed / Pew

Core Insights

Five findings that shape product, pricing, and channel strategy

T0 is the Largest Single Segment
27% of US adults have never used AI. They will not pay for "AI" — they will pay for a problem solved. Don't mention AI. Sell the result.
T2→T3 is the Biggest Growth Opportunity
22% are monthly free users. They don't lack intent — they lack a "worth paying for" scenario. Product positioning: bridge from free to paid.
T3 is the Paying Base
15% daily/weekly active. Willing to pay for productivity. Pricing anchor: save 1 hour = worth the monthly fee.
T4 is the Leverage Layer
4% influence 80% of industry discourse. They are the opinion source, OSS contributors, procurement advisors. Strategy: free/low-cost access to T4 → they pull T2-T3.
T Distribution Varies Wildly by Industry
No one-size-fits-all. Every new industry requires its own T-distribution assessment. The T-Model is a framework, not a template.

Where does your industry sit?

GWG Research applies the T-Model to every industry we analyze.
Find your market's T profile — and the right go-to-market strategy.

Electrical Decision Chain · GEO Report 2026
View Research Reports →