Before you build, check that someone wants it

Don't build on a hunch.
First, see what the market and real users are saying

AI reads Hacker News, Indie Hackers, Reddit and review sites, cross-checks competitor pricing and traffic, and writes a feasibility report where every claim has a source.

Real discussion · 25

“A wrong idea is expensive — it burns money, time, energy, and people.”

Indie Hackers · 2026-08

“Spending three months building everything first and then trying to go about validation.”

HN · 2025-12

“By the time the app was done, there were 3 competitors.”

HN · 2026-04

HNIHG2TP 4 platforms

Evidence grade Higher means harder evidence

AI AI inference Never used

HNIHG2TP 4 platforms cited

3. Landscape

3. Competitive landscape

AI validation reports are a crowded, low-traffic red ocean. Human-research platforms have the traffic and willingness to pay but only serve enterprises. Evidence-based validation for indie developers is still open.

Competitor	Category	Entry price	Reviews	Data credibility
Us	Human research	$49 / session	G2 4.6 (1,047)	High
Re	Human research	$40 / session	G2 4.6 (525)	High
Us	Human research	$30 / session	G2 4.9 (5)	Med-high
Lo	Ad smoke test	$29 / run + ad spend	None	Med-low
Id	AI validation report	Free / ~€20	Almost none	Low
Va	AI validation report	Free	Trustpilot 3.4 (1)	Low
Di	AI validation report	Free / $9	None	Low
Wo	AI report + user finding	Free / $5	None	Low

2. Verbatims

2. What users are saying

The real pain is not "not knowing to validate" but that it takes too long, target users are hard to find, and the information is untrustworthy. The 25 verbatims below come from comment sections, forums and review sites, with originals, translations and sources.

Theme	Count	Insight
A. Being wrong is expensive	3	Validation value is agreed on; price it by development time saved
B. Research and finding people take too long	6	Biggest pain: developers would rather code than research
C. Can't find target users	5	Cold outreach mostly fails, worse across markets
D. Can't trust the information	5	Polite interview answers and AI scores do not count
E. Only behavior and payment count	3	Paying is the highest validation signal
F. Unhappy with existing tools	3	Existing tools are pricey, have fake respondents, and doubtful value after paywalls

The hardest part is not the code, it is finding time to talk to potential users. The hardest part is not the code, it is finding time to talk to potential users.HN · 2026-06

Coverage

This report generates no opinions; it only collects and cross-checks evidence. The information is scattered across 15 sites and years of discussion that one person cannot gather systematically in a short time.

Evidence grade	Type	In this report
Grade A · behavioral	Target users pay or leave a deposit	Add human validation
Grade B · first-hand	Interviews with target users	Add human validation
Grade C · public voices	Spontaneous posts in communities and review sites	Part 2 · 25 items
Grade D · market facts	Pricing, traffic, ratings, funding	Parts 3–4 · 8 competitors
AI inference and scoring	—	Never used

greenlit.app / projects / report

Projects / New idea / Feasibility report

Greenlit idea feasibility report

Sample v3

Oct 9, 2026@TonyEvidence grade C · D25 verbatims8 competitors

Report ready6 / 6

Idea profileDone Collect verbatims25 Competitors and pricing8 competitors Traffic and fundingSimilarweb Grading and cross-checkC · D Conclusion and next steps31 citations

Coverage

Evidence grade	Type	In this report
Grade A · behavioral	Target users pay or leave a deposit	Add human validation
Grade B · first-hand	Interviews with target users	Add human validation
Grade C · public voices	Spontaneous posts in communities and review sites	Part 2 · 25 items
Grade D · market facts	Pricing, traffic, ratings, funding	Parts 3–4 · 8 competitors
AI inference and scoring	—	Never used

1. Your need

Item	Details
Your idea	A service that validates ideas for indie developers
Problem	Validating an idea takes weeks: scanning competitors, reading platform threads, finding people in the target market, writing it up
Target market	US, UK, Canada, Australia (the product market, regardless of where the user lives)
Deliverable	Feasibility report: verbatims, in-depth competitor profiles, conclusions and next steps
Target users	Time-poor indie developers · cross-market founders · first-time founders

2. What users are saying

Theme	Count	Insight
A. Being wrong is expensive	3	Validation value is agreed on; price it by development time saved
B. Research and finding people take too long	6	Biggest pain: developers would rather code than research
C. Can't find target users	5	Cold outreach mostly fails, worse across markets
D. Can't trust the information	5	Polite interview answers and AI scores do not count
E. Only behavior and payment count	3	Paying is the highest validation signal
F. Unhappy with existing tools	3	Existing tools are pricey, have fake respondents, and doubtful value after paywalls

The hardest part is not the code, it is finding time to talk to potential users. The hardest part is not the code, it is finding time to talk to potential users.HN · 2026-06

3. Competitive landscape

Competitor	Category	Entry price	Reviews	Data credibility
Us	Human research	$49 / session	G2 4.6 (1,047)	High
Re	Human research	$40 / session	G2 4.6 (525)	High
Us	Human research	$30 / session	G2 4.9 (5)	Med-high
Lo	Ad smoke test	$29 / run + ad spend	None	Med-low
Id	AI validation report	Free / ~€20	Almost none	Low
Va	AI validation report	Free	Trustpilot 3.4 (1)	Low
Di	AI validation report	Free / $9	None	Low
Wo	AI report + user finding	Free / $5	None	Low

4. Competitor deep dives

The 8 competitors fall into two groups: AI validation reports (competing with the report) and human-research platforms (competing with human validation).

IdAI validation report

Dimension	Details
Founded / entity	Milan-based company; site copyright since 2024
Positioning	"Validate in 120 seconds" AI suite: scoring, market size, business plan, branding and ad creatives
Vertical?	No: validation is the entry point; the focus is generating a full startup kit

Verbatims · 25 · 6 themes

A. Being wrong is expensive 3

B. Research and finding people take too long 6

C. Can't find target users 5

D. Can't trust the information 5

E. Only behavior and payment count 3

F. Unhappy with existing tools 3

HNIHG2TP 4 platforms

Competitor traffic · last 3 months

Human-research platforms get the most traffic

Us $49 / session

Re $40 / session

Us $30 / session

Too little traffic, no data

Lo $29 / run + ad spend

Too little traffic, no data

SW Source: Similarweb

data source sites

sourced verbatims

full competitor profiles

target markets: US UK CA AU

A–D

evidence grades, labelled per claim

10 min

from submission to report

New AI capability → what to build

What AI just learned, and what you could build with it.

Each new capability comes with a few directions. Pick one and take it to validation to see if anyone wants it.

Demo data

Voice

New capability

Realtime voice APIs are now fast enough for natural conversation, interruptions included

Directions to build

A meeting app that coaches how you speak and ask questions, live

For: Online educators Validate this

A voice agent that small shops can connect to their phone line in minutes

For: Small local businesses Validate this

Vendor release notes · this week

Agents

New capability

Models can chain dozens of tool calls to finish a task instead of answering once

Directions to build

A small tool that tracks competitors daily and sends a brief

For: Indie developers and small teams Validate this

Record a repetitive back-office task once, then let it run itself

For: Ops and finance roles Validate this

Hacker News · this week

Multimodal

New capability

Models can read screenshots and hand sketches and suggest what to do next

Directions to build

A tutor that watches your canvas and teaches drawing step by step

For: Hobbyist beginners Validate this

Upload a UI screenshot and get accessibility and copy checks

For: Frontend developers and designers Validate this

GitHub Trending · this week

Capabilities come from public sources. The directions are editorial interpretation, published after human review. They are inspiration, not proof of demand.

Startup market

The startup market: lots of money, concentrated in a few companies.

Four signals for people about to build: where the money went, how concentrated it is, early-stage funding, and developer numbers. Every number has a source.

Capital $212B

Global AI venture funding in 2025, +85% YoY

What it means for foundersMoney is concentrating in AI; riding the niche makes you easier to notice.

Concentration ~65%

of global venture funding went to just four companies in Q1 2026

What it means for foundersBig companies absorb the capital; small teams win on narrow needs.

Early stage $4.9B

Seed and angel rounds in Q2 2026, −15% QoQ, −27% YoY

What it means for foundersEarly funding is tighter; validate cheaply before committing.

Developers +178%

YoY growth in new repos using an LLM SDK

What it means for foundersMore people are building AI products; competition gets more crowded.

AI funding grew almost 5x in four years

Real data · Crunchbase

Global AI venture funding, $B. H1 2026 is an estimate (dashed).

AI share of global venture: 34% to about 80%

Real data · Crunchbase

Share of global venture funding going to AI companies.

Where the money went

Real data · Crunchbase

Destination of global AI venture funding in 2025.

By region

Bay Area 60%Rest of US 19%Other 21%

By company type

Foundation models ~40%Other AI companies ~60%

Funding is concentrated in the Bay Area and foundation-model companies. Ordinary teams get less, so validating demand first matters more.

Four more numbers on AI funding in 2025

Real data · Stanford / Crunchbase

+71%AI companies newly funded

15 → 28Billion-dollar rounds

+200%Generative AI funding growth

$126BGlobal AI funding raised in the Bay Area alone

Real revenue of popular software

Sample data showing the board layout. Real verified revenue appears here once TrustMRR is connected and permitted.

Demo data

SoftwareVerified MRRMoM

1 Sa

Sample A

$2.40M/ mo ▲ 6%

2 Sa

Sample B

$700K/ mo ▲ 12%

3 Sa

Sample C

$380K/ mo —

4 Sa

Sample D

$240K/ mo ▲ 8%

5 Sa

Sample E

$210K/ mo ▲ 20%

6 Sa

Sample F

$190K/ mo ▲ 1%

7 Sa

Sample G

$170K/ mo —

How to read this board

Heavily top-loaded

The #1 earns about 3.4x the #2; the top 7 total about $4.29M / month.

The rest sit much closer

Ranks 3 to 7 sit between $170K and $380K a month, a scale small teams reach.

Some are for sale

Listed software can also be an acquisition target; the board marks "for sale", and price and multiple are on the source site.

New launches

Data from Hacker News Show HN. See what others just shipped.

Synced · 24 minutes ago Real data · Hacker News (Show HN)

Ll Llmspeed

feel how fast a local LLM runs before you buy the hardware

▲ 328 86 3 days ago

Sp Sphinx

A hook that quizzes you on your AI's code before shipping

▲ 252 89 2 days ago

le legendofhyrule.com

The Zelda timeline to catch up before the Ocarina of Time remake

▲ 213 125 yesterday

Es Esp32sim

Run ESP32 firmware in the browser

Open source

▲ 205 46 2 days ago

ce celphase.github.io

High-quality duck audio synthesizer

▲ 175 44 2 days ago

wh whatsapp-mcp

WhatsApp MCP running free on Cloudflare

AIOpen source

▲ 146 78 yesterday

Data from Hacker News · Show HN

A hot market does not mean your idea has demand.

Greenlit shows what real users complain about in your niche and what competitors charge.

Validate my idea

Sources: Crunchbase 2025 Crunchbase 2026 Q2 Silicon Canals / Crunchbase 2026 Q1 Stanford AI Index 2026 GitHub Octoverse 2025。2022–2023 figures and the Q2 2026 seed figure are quoted from media coverage of Crunchbase.

What's inside

One report, four kinds of evidence, all traceable.

The real structure of the sample report. Click to switch sections.

data source sites 15

sourced verbatims 25

full competitor profiles 8

independent reviews cited 1,578

verbatim time span 2020–2026

AI inferences or scores 0

25 verbatims

By demand theme

Being wrong is expensive 3

Research and finding people take too long 6

Can't find target users 5

Can't trust the information 5

Only behavior and payment count 3

Unhappy with existing tools 3

The biggest theme is "research and finding people takes too long" at 24% of all verbatims, followed by "can't find target users" and "can't trust the information", 20% each.

Verbatim sources

4 sites

Hacker News 14

Indie Hackers 8

G2 2

Trustpilot 1

Verbatim year

22 dated

4 2020

1 2022

3 2025

14 2026

3 Undated

8 competitors · category

By positioning

Human research 3

AI validation report 3

Ad smoke test 1

AI report + user finding 1

Competitor data credibility

Independent reviews and scale

High 2

Med-high 1

Med-low 1

Low 4

Data sources

15 sites in total, some below

Communities

Hacker NewsIndie Hackers

Review sites

G2Trustpilot

Market data

Pricing pagesSimilarwebCompany database

Evidence grades

Every claim shows how hard its evidence is.

No AI scoring, only collected and cross-checked evidence. You can see at a glance what is fact and what still needs validating.

Behavioral

Target users pay or leave a deposit. The strongest signal.

Add-on

First-hand

Interviews with target users.

Add-on

Public voices

Spontaneous posts by users in communities and review sites.

Included · 25

Market facts

Official pricing, traffic, ratings and funding.

Included · 8

No AI inference or scoring anywhere in the report. AI collects, translates and organizes. Conclusions come only from the four evidence types above.

How it works

In ten minutes, from one sentence to a report you can decide on.

Step 1

Describe your idea

A few sentences is enough. We turn it into a structured idea profile and ask a few clarifying questions if needed.

Step 2

Agents collect in parallel

Verbatims across platforms, competitor sites and pricing, traffic and funding data are collected and cross-checked at the same time.

Step 3

Read the report, add validation

Get a report with citations. Want grade A or B evidence? Add human validation.

From $9.9 per report · or subscribe for ongoing monitoring

Before you build, get an answer with sources.

If we cannot complete your report, the validation is returned automatically.

Try it free See a sample report

Don't build on a hunch.First, see what the market and real users are saying

Greenlit idea feasibility report

Coverage

1. Your need

2. What users are saying

3. Competitive landscape

4. Competitor deep dives

Human-research platforms get the most traffic

What AI just learned, and what you could build with it.

The startup market: lots of money, concentrated in a few companies.

AI funding grew almost 5x in four years

AI share of global venture: 34% to about 80%

Where the money went

Four more numbers on AI funding in 2025

Real revenue of popular software

How to read this board

New launches

A hot market does not mean your idea has demand.

One report, four kinds of evidence, all traceable.

25 verbatims

Verbatim sources

Verbatim year

8 competitors · category

Competitor data credibility

Data sources

Every claim shows how hard its evidence is.

Behavioral

First-hand

Public voices

Market facts

In ten minutes, from one sentence to a report you can decide on.

Describe your idea

Agents collect in parallel

Read the report, add validation

Before you build, get an answer with sources.

Don't build on a hunch.
First, see what the market and real users are saying