Insights · Launch database

What 30 one-person and tiny AI-era businesses
have in common — and where they split.

Patterns that show up only when you read every case at once. Hover any chart to see which cases the bar contains.

87/100
Median fit score
9
Median sources
8
Median research depth
18
Flagship teardowns
12
Structured briefs
43%
Solo / full-time signal
01

The shape of the fit-score distribution.

The highest-fit cases combine observable distribution, concrete monetization and a model a solo founder can actually study.

< 75 fit
4
13%
75-84 fit
7
23%
85-91 fit
12
40%
92+ fit
7
23%
02

Which business models look most reusable.

Paid databases, SaaS and education products rank well because their offer, channel and operating model are easier to observe from public sources.

LABEL
Cases
Avg fit
Paid Database
1
98/100
AI Output Product
1
96/100
Digital Product
1
96/100
Agency
1
94/100
Founder Interview Database
1
90/100
AI Tools Directory
2
89/100
Trend Intelligence Database
1
87/100
Education
7
84/100
Curated Inspiration Directory
1
84/100
SaaS
9
82/100
Newsletter
5
82/100
03

Where distribution is most visible.

X and SEO are the clearest public surfaces in this dataset. X exposes the founder's cadence; SEO exposes the product's demand capture.

LABEL
Cases
Avg fit
X
8
90/100
SEO
5
89/100
Newsletter
4
80/100
High Converting Official Lan
1
94/100
LinkedIn
1
94/100
Substack
1
91/100
Product-led
1
90/100
YouTube
1
88/100
Community
1
86/100
Content
1
86/100
Free Technical Blog With Dee
1
82/100
Developer ecosystem
1
75/100
Word of mouth
1
75/100
Referral Program With Leader
1
72/100
Positioning
1
72/100
Launch platforms
1
70/100
04

The actual AI leverage pattern.

AI-native and AI-assisted are different cases: one sells AI output, the other uses AI to lower research, support or production cost.

AI-era reference16
avg fit 80/100
AI-assisted5
avg fit 90/100
AI leverage5
avg fit 92/100
AI media4
avg fit 89/100
AI-native2
avg fit 97/100
05

Source depth vs OnePersonAI fit.

The best cases are not just famous. They have enough public evidence to make the operating pattern readable.

0/10026/10051/10077/100103/100024681012Cited source countOnePersonAI fit scorePieter LevelsDanny PostmaMarc LouTony DinhDamon ChenBrett WilliamsJustin WelshLenny RachitskyRowan CheungAJPat WallsAndrei / There's An AI For ThatMatt WolfeDan KoeBen TossellBrian Dean and Josh HowarthDaniel VassalloArvid KahlWes BosRob HopeJosh ComeauJon YongfookNicolas Cole and Dickie BushTiago ForteMike PerhamMarie Martens and Filip MinevShaan Puri & Ben LevyPaul Jarvis & Jack EllisJustin DukeDan Ni
channeldatabrandtechspeedniche
06

Which moat types show up most often.

Brand and channel moats are easier to see publicly. Data and tech moats require stronger source notes before we overclaim.

LABEL
Cases
Avg fit
niche
1
91/100
brand
8
87/100
data
6
85/100
tech
11
84/100
speed
4
84/100
07

Who actually builds these.

The default story is 'an engineer who learned marketing,' but that's only ~40% of the dataset. Designers, marketers and operators each have a real lane.

Solo13
Small9
Founder-led8
08

How source confidence is distributed.

This launch version does not pretend every case has the same evidence quality. The useful view is confidence and depth, not fake geography.

LABEL
Cases
Avg fit
A/B
23
86/100
A
6
85/100
B
1
72/100
09

Research depth across cases.

Median research depth = 8. Range: 8-10. Lower-depth cases stay useful for comparison, but should not be over-positioned as premium-grade teardowns yet.

0-2
2-4
4-6
6-8
28
8-10
2
10-12
12-14
14-16
16-18
18-20
wk

Each insight links to the cases behind it.

Members can filter the database by any cell, inspect the cases behind each pattern, and follow revisions as source confidence improves.