30 Days of Product · Workshop
Workshop · 90 minFor tech-savvy builders
Idea
to App.
Playbook to build and make money from your app idea.
Your idea
Problem worth solving
Pick a real problem. Validate fast. Ship the loop. Charge for it.
by you · solo · in months
First $
$1k+
on the way to $100k
Stripe
New subscription · $29/mo · Customer #1
Just now
Idea
Validate
Ship
Pre-sell
Waitlist
Beta
Adrian Cole
Founder · 30 Days of Product
Who's running this
Hi — I'm Adrian.
Adrian Cole
Product Strategist · Founder · Builder
30+
Products shipped
10+
Years in product
3
Startups built
My story · in 4 lines
10 years shipping products — hardware to software to AI agents. I've worked at early-stage startups and ran product at a $100M company while it was scaling.
I've built 3 startups from zero. Some won, some didn't. Every one taught me something I'm passing to you today.
My job today: make sure you walk out of this room ready to ship.
By the end of today
You'll walk away with this.
Not theory. Not "someday." Today. The four things you need to actually start.
01
A real problem
worth solving
worth solving
Not a feature. Not a vibe. A specific pain you can describe in one sentence.
02
A clear
validation plan
validation plan
5 customer conversations. The right questions. The signal to listen for before you write a line of code.
03
Your App
Blueprint
Blueprint
User · Problem · Outcome · 5 features · Primary workflow. One page. AI-ready.
04
A path to
your first $1k
your first $1k
Pricing model, channel, and first-10-customers playbook. From idea to revenue.
Momentum beats perfection. Every time.
Section one · why now
The big
opportunity now.
The barrier to launching has collapsed. AI didn't just change what's possible — it changed what's cheap.
01
Then → now
Capital · gone
$500k seed → $20/mo in tools
02
Then → now
Team · optional
5–10 people → solo
03
Then → now
Time · weeks
6 months → 2–4 weeks
The north star
The barrier is gone.
You don't need permission. You don't need capital. You don't need a team. You need a problem worth solving, a loop to ship it, and the guts to charge for it. The tools are already on your laptop.
Capital
removed
$20/mo subscription replaces $100k seed round.
Team
optional
Solo founders are shipping what used to take 5 people.
Distribution
DIY
Twitter · LinkedIn · Reddit · Product Hunt — all free.
The challenge now is judgment + strategy.
What changed · the math
Old way vs. new way.
2015 · OLD MODEL
Idea → Spec → Dev
→ Months → First Look
→ Months → First Look
Idea
→ Hire engineer
→ Write spec
→ Wait 3–6 months
→ See first interface
→ Hire engineer
→ Write spec
→ Wait 3–6 months
→ See first interface
Money to start
$50k+
Time to first look
3–6 mo
2026 · NEW MODEL
Idea → Blueprint → AI
→ Same-day Feedback
→ Same-day Feedback
Idea
→ Structured thinking
→ AI prototype
→ Show real people
→ Same-day feedback
→ Structured thinking
→ AI prototype
→ Show real people
→ Same-day feedback
Money to start
$20/mo
Time to first look
1 day
5 years ago you needed engineering to see your idea. Now you need clarity.
Under the hood
What the tech unlocked.
Four shifts collapsed the gap between "idea" and "real thing you can show someone." None of these existed at this quality 3 years ago.
LLMs
understand intent
understand intent
Describe what you want in English. The model gets it. No spec doc required.
AI generates
working UI
working UI
Lovable, v0, Base44, Bolt — prompt to clickable interface in minutes, not weeks.
Backend is
abstracted
abstracted
Supabase, Firebase, Stripe — auth, database, payments. Plug it in, ship it.
Hosting is
one click
one click
Vercel, Netlify, Railway — push code, get a live URL. No DevOps. No servers.
You describe what you want. The stack builds it. Friction is being removed.
Predict · then reveal
They couldn't code. And needed millions.
Every one of these founders had two giant barriers: they couldn't write production code, and they had to raise millions to launch. AI killed the first barrier. The funding game changed too. Hover each card to see what they built.
Founder · est. 2009
Jennifer
Hyman
Hyman
Harvard MBA · hospitality background · no engineering
$540M raised · IPO 2021
She built →
Rent the
Runway
Runway
Designer dresses by mail — the original "subscribe to your closet." Built a rental marketplace where Net-a-Porter built retail.
Rental marketplace · $540M+ total raised
Founder · est. 2014
Whitney
Wolfe Herd
Wolfe Herd
Marketing + business background · ex-Tinder co-founder · no code
$424M raised · IPO 2021 · $1B+ valuation
She built →
Bumble
A dating app where women message first. Female-led friction redesign of the dating market.
Dating + social · $424M raised · $1B+ valuation
Co-founder · est. 2008
Brian
Chesky
Chesky
Industrial designer at RISD · sold cereal boxes to fund the first round
$6B+ raised · IPO 2020 · $80B market cap
He built →
Airbnb
A marketplace for renting people's spare rooms. Hospitality without hotels. Designed for trust before tech.
Travel marketplace · $6B+ raised · $80B market cap
Founder · est. 2011
Katrina
Lake
Lake
Economics + consulting · started from Harvard Business School project
$42M raised · IPO 2017 · $2B valuation
She built →
Stitch Fix
Personal stylist by algorithm + mail. Curated boxes — you keep what you want. Retail meets recommendation engine.
Fashion subscription · $42M raised · IPO 2017
Founder · est. 2013
Ryan
Hoover
Hoover
PM at a gaming startup · started as an email newsletter
$8M raised · acquired by AngelList · $20M+
He built →
Product
Hunt
Hunt
A daily list of new tech products, voted on by the community. The discovery layer for indie launches.
Community + discovery · $8M raised · acquired $20M+
They couldn't code and needed millions to ship. You need neither. That's what changed.
Quick quiz
The catch
What tools won't do.
AI removed the building cost. It did not remove the thinking cost. The founders who win this era will be the ones who pick problems worth solving — and have the discipline to validate before they build.
Tools won't
Pick your
problem
problem
You need to know what's broken — and for whom.
Tools won't
Talk to your
customers
customers
Validation is a human conversation. AI can simulate users; only real ones tell you the truth.
Tools won't
Charge
for you
for you
Pricing, channels, sales — still on you. The good news: easier than ever.
Tools 10x your speed. They don't 10x your judgment.
Then · vs · now
What it took to launch.
Same idea. Same first product. Two different eras of what it actually cost to ship a v1 into customers' hands.
2015 · the old way
A team + a seed round
👥
People needed
3–8 (founders + engineers + designer)
💰
Funding required
$500k – $2M seed round
⏱
Time to first prototype
6–12 months
🎯
Time to first paying customer
12+ months
2026 · the new way
You + a $20 subscription
👤
People needed
1 — you, solo
💰
Funding required
$20 – $200 / month in tools
⏱
Time to first prototype
1 day
🎯
Time to first paying customer
2–4 weeks
Capital floor dropped 1,000×. Team floor dropped to one. The only thing that didn't drop: your judgment.
Section two · what to build
Where to look
for ideas.
Not every app idea is created equal. Three categories where solo builders actually win — and one is probably staring you in the face.
01
Path
Personal product
Build the thing you wish existed
02
Path
Internal tool
Build for your team or company
03
Path
Vertical SaaS
Build for a niche you know
Three paths
Pick your path.
Different contexts. Different risk tolerance. All three are real businesses. Most successful first-time founders pick lane 1 or lane 2 — and graduate to lane 3.
01
Lowest risk · fastest learning
Personal
Products
Products
Build the thing you wish existed. You are the user. You feel the pain. You ship fast.
- Fitness dashboard
- Habit tracker
- Reading queue
- Personal CRM
if it works for you, it works for 10,000 like you.
02
Captive audience · paid problem
Internal
Tools
Tools
Build for your company or the teams you work with. Internal users don't need polish — they need speed.
- Marketing ops dashboard
- Sales pipeline tracker
- Onboarding automation
- Compliance checklist
internal wins become external products.
03
Biggest upside · highest risk
New Product
Exploration
Exploration
Build for a market you don't yet serve. Higher reward, harder validation. Demands talking to customers.
- SMB SaaS niche
- Vertical AI agent
- Marketplace
- Consumer app
the path most "real startups" take.
Lane 01 · personal products
Build for yourself. First.
Example · personal fitness dashboard
Fitness OS.
A single dashboard for your workouts, nutrition, bloodwork, and progress photos — built around how you actually train.
Track workouts + nutrition in one place
Sync wearables (Whoop, Oura, Apple Health)
Log bloodwork over time
Weekly progress photo timeline
Goal-based dashboards (cut · bulk · maintain)
Why this lane wins
- You are the user — feedback loop is instant
- Low security risk — only your data
- Failure is cheap — you learn either way
- Built-in audience: people exactly like you
Money path
$5–15/mo
Consumer subscription. Niche audience of 2k–10k users = $10k–100k MRR.
Lane 02 · internal tools
Build for your team. Or your boss.
Example · revenue ops dashboard
Pipeline Pulse.
A live revenue dashboard for the sales team — pipeline stages, campaign performance, attribution, and weekly forecast in one screen.
Sync with HubSpot · Salesforce · Stripe
Pipeline stages with deal-level drill-down
Campaign performance by channel
Weekly forecast vs target
Slack alerts for stalled deals
Why this lane wins
- Captive audience — they already work with you
- Need clarity + speed, not polish
- Pain is obvious + measurable
- Often becomes a paid SaaS for other teams
The math · internal SaaS
50
×
$500/mo
= $25k MRR · $300k ARR
50 companies. Mid-market SaaS. Solo founder. Done inside 18 months.
Path 03 · new product exploration
Build for a market you already know.
Example · SaaS for a niche vertical
AgencyOS.
Operations SaaS for boutique creative agencies (10–30 person teams). Project pipeline + retainer billing + client portal in one tool. Niche enough to dominate, sticky enough to retain.
Project pipeline + status board
Retainer + recurring invoice automation
Client portal · deliverable approval
Time + budget tracking per project
Team workload + utilization view
Why this lane is hardest
- You're not the user — must talk to real ones
- Distribution = cold outreach + niche communities
- Higher reward · longer validation runway
- Your unfair advantage: pick a vertical you already work in
The math · niche vertical SaaS
100
×
$250/mo
= $25k MRR · $300k ARR
100 niche customers. $50–$250 / mo range. Best edge: pick a vertical you already know.
The fastest founders pick a vertical they already understand. Specialized knowledge = unfair advantage.
Section three · the idea
Getting
good ideas.
Most app ideas die in the first 30 seconds — because they're a feature, not a problem. Let's fix that.
01
Source
Pain-led
"I keep running into this"
02
Source
Boring industry
Still run on spreadsheets
03
Source
Wedge
Big tool does X badly
The mindset reset
Bad ideas come from wanting to build.
Good ideas come from noticing. The best founder skill in this era isn't coding — it's pattern recognition on friction.
Bad source
"I want to build
an app that…"
an app that…"
Starts with the building. The idea is whatever fits the urge. No user. No pain. No why. These die in 3 months — if they even ship.
"Uber but for…" "It's like Airbnb but…"
Good source
"I keep running
into this…"
into this…"
Starts with the friction. You noticed something is broken. You can name who feels it. You can describe what it costs them. The build comes after.
"This shouldn't be this hard."
Carry a notes file. Write down every time something pisses you off this week.
Four shapes of good ideas
The four idea archetypes.
Almost every solo-founder win fits one of these. Knowing the shape helps you pressure-test your own.
01
Pain-led
idea
idea
You felt the pain personally. You can describe it in one sentence. The market has 10k+ people like you.
"I waste 2 hours every Sunday planning my week."
02
Boring-
industry
idea
industry
idea
A real workflow in a non-sexy industry is still run on spreadsheets + email. You replace the spreadsheet.
"Dental offices still send reminders by hand."
03
Wedge /
unbundle
unbundle
A big tool does 30 things badly. You do one of them 10× better. Steal one wedge of a giant.
"Notion sucks at recurring tasks. I'll fix only that."
04
Trend-
leverage
idea
leverage
idea
A new capability (AI agents, voice, on-device LLMs) makes a previously-impossible product possible. You go first.
"AI agents are now reliable enough to run my inbox."
The strongest archetype
Pain-led ideas win the most.
5 tests for a pain-led idea
01
Can you name a real person who has this pain? (Not "people who…")
02
Can you describe the pain in one sentence a 12-year-old would understand?
03
Are they already paying something to solve it badly? (spreadsheet, VA, three tools)
04
Does it happen weekly or more often? Frequency makes products sticky.
05
If you fixed it, would they actually pay $10/mo to make it go away?
Real-world example
Every Sunday I spend 2 hours figuring out what I'm doing this week — across 4 tools.
— solo founder, 30 yrs, fitness coach
Real person · weekly pain · already pays for Notion + Calendar + Notes + Apple Reminders.
Now ask: If a single tool replaced all four and saved her 90 minutes a week — would she pay $15/mo? Almost certainly yes.
That's a pain-led idea.
Now ask: If a single tool replaced all four and saved her 90 minutes a week — would she pay $15/mo? Almost certainly yes.
That's a pain-led idea.
Predict · then reveal · round 1
Which idea has legs?
Two pitches. Same founder. Same effort. One will get traction. Hover each card to find out which — and why.
Idea A.
"An AI app that helps people get fit."
Personalized workouts. Nutrition tracking. Motivational quotes. Social feed. Streak gamification. Maybe a marketplace for coaches.
Target user: "everyone who wants to be healthy"
crowded market
no specific user
5+ features
Dies in 3 months
No specific user. No specific pain. Competing with MyFitnessPal, Whoop, Strava, Apple Fitness — all 100M+ users with deep pockets.
Five features means none are great. You can't win on "fitness."
Idea B.
"A workout-partner matcher for 25–40-year-old pros in major cities."
Match by schedule + experience level + neighborhood. One core feature. One workflow. One very specific user. Pay $9/mo for guaranteed matches.
Target user: 28-yr-old Sarah · lives in Brooklyn · lifts 4×/week · hates training alone
specific user
weekly pain
one feature
This one ships
A real person you can find on LinkedIn. A weekly pain. A single workflow. No competitor doing exactly this.
10,000 users × $9/mo = $1M/year. Solo. From a phone.
Quick quiz
Predict · then reveal · round 2
Round two. Same question.
Both pitched to me this year by real founders. One got to $5k MRR in 90 days. Hover each card to see which.
Idea A.
"Reminder + intake software for independent dental offices."
Replaces the front-desk's spreadsheet + email chain. Saves 6 hrs/week of admin. $200/mo per office. Sold 1 at a time via cold outreach.
Target user: Mike · owns 2 dental practices · pays a part-time admin $1,200/mo
boring industry
spreadsheet replacement
pays $200/mo
$5k MRR in 90 days
Real workflow. Real spreadsheet. Real money already spent on staff time. Pitch sells itself: "replace 6 hours/week of front-desk admin."
25 dental offices × $200/mo = $5k MRR. No marketing. Cold outreach + 1 case study.
Idea B.
"A Web3 social network for creators with NFT monetization."
Disrupting Patreon, Substack, and OnlyFans. Decentralized creator economy. Token-gated content. Community-owned. Going to change everything.
Target user: "creators who want to own their audience" — couldn't name a single one
disrupting 3 giants
buzzword stack
no first user
Never shipped
Disrupting three incumbents at once. Trend-stacking (Web3 + NFTs + social + creators). Could not name a single creator who'd switch from Patreon today.
Cool deck. Zero distribution path. Zero willing first customer. Big ambition ≠ good idea.
Quick quiz
Steal this · for paths 2 + 3 (Internal · Vertical SaaS)
Your AI idea generator.
This template targets Internal-Tool and Niche-Vertical SaaS ideas — lanes 2 and 3. It mines your work history, communities, and daily tools to surface ideas you have an unfair edge on.
Stage · idea generation
You are a startup advisor who has reviewed 1,000+ solo-founder products. Help me generate 10 pain-led app ideas — biased toward Internal Tools, Niche-Vertical SaaS, and New-Product Exploration (B2B-leaning) — that I could realistically ship in 60 days.
About me:
- Background: [your job / industry]
- Communities I'm part of: [Slack groups · subreddits · LinkedIn niches]
- Tools I use daily: [apps · spreadsheets · workflows]
- Pain points I keep running into: [list 3–5 frictions]
For each of the 10 ideas, give me:
1. One-sentence pitch (who · pain · solution)
2. The user I should interview first — their job + how to find 5 of them
3. What they already do to (badly) solve this today
4. Pricing model + realistic price point
5. Risk score (1–5) — likelihood I can validate in 2 weeks
Bias toward boring industries · captive audiences · weekly pain. Avoid "disrupt X" framings. Avoid ideas that require a network effect to be useful.
End by ranking the 10 by likelihood of $1k MRR in 90 days.
Run this once a week. Keep a running list. Most ideas die — that's the point.
Section four · the work most founders skip
Validation
before code.
The #1 way founders lose money in 2026: they ship a beautiful AI-built app for a problem nobody has. Let's not do that.
01
Move
Talk to 5
5 customer interviews · strangers
02
Move
Find signal
Past-tense behavior · not opinions
03
Move
Pre-sell
$1 paid = real proof
The most expensive mistake
Founders build. Then ask.
The order is wrong. In a world where building is cheap, the bottleneck is knowing what's worth building. Most never check.
The number that matters
42%
of failed startups cite "no market need" as the #1 reason they died. Not tech. Not money. Demand they never validated.
— CB Insights · "Top Reasons Startups Fail"
01
Build for 2–6 weeks based on a hunch
02
Launch on Product Hunt · get 12 signups
03
Realize no one's using it a month later
04
Quietly abandon · blame distribution
05
Start a new project. Repeat the loop.
Validation isn't a phase. It's a discipline.
Anatomy · click each step
The validation ladder.
5 steps · from idea to first $
01
Pain interview
+
02
User clarity
+
03
Solution check
+
04
Pre-sell / waitlist
+
05
Beta + first $
+
01
Pain interview
Talk to 5 real people who match your target user. Don't pitch. Ask what they currently do to solve this pain, how often they hit it, and what they've already tried.
budget · 5 days · ~$0
Click each step
Steal this · use today
Customer interview script.
Paste this into Claude or ChatGPT. Edit the [brackets]. You'll have a tailored interview script for your idea in 30 seconds.
Stage · customer interviews
You are a senior product researcher running a discovery interview. Help me prepare a 30-minute interview script for the following idea.
Idea: [one-sentence problem + user]
Person I'm interviewing: [their job / context / how I found them]
Generate a script with:
1. Warm-up (3 min) — questions that get them comfortable, no pitching
2. Pain discovery (10 min) — 5 open questions that surface the problem from their mouth, not mine. Use "tell me about the last time…" framing.
3. Current solution (8 min) — what do they do today? What tools? How long does it take? What does it cost (time or money)?
4. Willingness to pay (5 min) — frame as: "if a product existed that did X, would you pay for it?" then probe price sensitivity.
5. Close (4 min) — would they look at a beta? Can I follow up in 2 weeks? Who else should I talk to?
Rules:
- Don't ask leading questions. Never say "wouldn't it be great if…"
- Don't pitch your solution until step 4.
- Listen for emotional language ("I hate it when…", "I waste hours…", "I'd kill for…") — that's signal.
End with 5 follow-up question variations if they go quiet.
5 interviews this week. Not 50. Not "later." 5. By Sunday.
Predict · then reveal
Signal or noise?
You interview a potential customer. They say one of these. Which is real signal — and which is polite encouragement that means nothing? Hover each card.
Yeah, that sounds cool. I'd probably use that.
— prospect, after a 20-min pitch
Noise · be polite
"Probably" means no. Future-tense politeness. They're being nice. They've signed up for zero things. Discard.
I love it. You should totally build this.
— friend over coffee
Noise · friend tax
Friends + family lie to protect you. Enthusiasm with no commitment is worthless. "Would you pay $20 right now?" cuts through.
I built a hacky Zapier workflow last month to do exactly this. Can I see it Tuesday?
— prospect, unprompted
Signal · gold
Past-tense action. They've already paid in time. They're pulling, not being pushed. This is your first customer.
Send me an invoice. If it saves me 5 hrs/week, $200/mo is a steal.
— SMB owner, ops director
Signal · pre-sale
Money on the table before the product exists. Pre-orders are the strongest validation you can get. Take it.
Quick quiz
Don't fool yourself
Validation false positives.
Four things that feel like validation but aren't. If you're doing any of these, you're lying to yourself.
False positive 01
Asking friends + family
They love you. They'll lie. Their "great idea!" means nothing. Strangers tell the truth.
do this → cold outreach to strangers
False positive 02
Leading questions
"Wouldn't it be cool if…" guarantees a yes. You learn nothing. Ask about past behavior, not future intent.
do this → "tell me about the last time…"
False positive 03
Twitter likes
A viral tweet is not a customer. Email signups are not customers. Even waitlists aren't customers. Charging is.
do this → ask for a $1 pre-order
False positive 04
Over-validating
50 interviews and no shipped product = procrastination. After 5 strong signals, stop researching. Build.
do this → 5 interviews → ship
Section five · structured thinking
The App
Blueprint.
One page. Five sections. The bridge between "I think I have an idea" and "here's what I'm building." This is the artifact AI needs.
01
Block
User
↓
02
Block
Problem
↓
03
Block
Outcome
↓
04
Block
5 Features
↓
05
Block
Primary workflow
Block
User
Who, specifically
02
Block
Outcome
What changes for them
03
Block
5 features + flow
The minimum set