Test willingness to pay before you build
By Tubemk Studio · Sold by Tubemk
Product R&D for creators with an audience: test willingness to pay before you build.
Free
Works with the AI agents you already use
ClaudeChatGPT
By Tubemk Studio · Sold by Tubemk
Product R&D for creators with an audience: test willingness to pay before you build.
Free
Works with the AI agents you already use
Tubemk MCP lets your AI find and fetch skills. Running a local skill still needs the install step below.
Connect Tubemk to Claude
RecommendedFree, read-only, no sign-in. Then ask Claude: "Use the Tubemk skill creator-product-rnd"
Not working? See the setup guide

git clone https://github.com/wanghie/tubemarket-skills tubemarket-skills cd tubemarket-skills python3 -m venv .venv && source .venv/bin/activate pip install -r requirements.txt pip install mediapipe opencv-python ./setup
---
name: creator-product-rnd
description: An R&D process for a new product by a creator/founder who already has an audience, shared across 6 formats (physical/digital/SaaS/service/course/community). Forces the agent to validate willingness-to-pay before building, prototype as cheaply as the format allows, and measure retention (Sean Ellis) before scaling. Runs inside Claude or ChatGPT, no API key. Triggers on "creator product R&D", "build a new product", "what product should I sell", "prototype this idea", "what should a creator sell".
---
# creator-product-rnd
A product lives or dies BEFORE it is built, not after. A 7-step framework that forces each checkpoint before effort is spent.
## Evidence this framework is NOT made up
Distilled from 6 primary-source frameworks with real, verified cases:
| Framework | Author | Source | Case + numbers |
|---|---|---|---|
| Sales Safari (30x500) | Amy Hoy | stackingthebricks.com | Noko/Freckle Time — bootstrapped by Amy, $38k MRR |
| Value Equation ($100M Offers) | Alex Hormozi | acquisition.com | Gym Launch — $17M profit in year 1, $46M exit |
| Ladders of Wealth Creation | Nathan Barry | nathanbarry.com/wealth-creation | Kit (ConvertKit) $43.8M ARR, bootstrapped |
| PMF Survey "very disappointed ≥40%" | Sean Ellis | pmfsurvey.com | Superhuman — 22% → 58% (First Round Review case) |
| Do Things That Don't Scale | Paul Graham | paulgraham.com/ds.html | Airbnb (photographing apartments by hand), Stripe (Collison installation) |
| Cohort-Based Course | Wes Kao | weskao.com, maven.com | altMBA 25k alumni; Ali Abdaal PTYA $1.9M/cohort |
## Frameworks this DROPS (with reasons)
- **Kevin Kelly "1000 True Fans"** — kept as north-star math (1000 × $100 = $100k/year), NOT used as a playbook. Kelly gives no concrete proving case; the "success cases" attached to it are retrofits.
- **Kickstarter "Coolest Cooler" $13M** — never quoted as a success case; it went bankrupt in 2019 without delivering. Use Pebble Time $20M or Peak Design Everyday Messenger $6.57M instead.
- **"MegaMaker pre-sale $50k"** — no traceable primary source. Use Justin Jackson's "Marketing for Developers, 5,000+ copies" instead.
## When to use
Before spending >2 weeks building a new product for a creator/founder with an audience. Skip the framework if:
- The product takes 1 week (a $9 mini digital product)
- You have no audience yet (≥1,000 engaged emails/followers is the minimum)
## 6 product formats — pick 1 before starting
| Format | Examples | Margin | Moat | v0 build time |
|---|---|---|---|---|
| Physical | Feastables, Chamberlain Coffee | 20-40% | Distribution + brand | 3-12 months |
| Digital (one-time) | Notion template, ebook, preset | 90%+ | Curation + trust | 2-6 weeks |
| SaaS | Kit, Beehiiv | 70-85% | Product depth + switching cost | 3-9 months |
| Service | Agency, 1:1 coaching | 60-80% | Reputation + track record | 1-2 weeks |
| Course | PTYA, B-School, Creator MBA | 90%+ | Brand + community | 4-12 weeks |
| Community | Skool group, Circle mastermind | 80-95% | Network effect | 2-4 weeks |
Picking a format = trading off margin / moat / time / personal fit. Do not pick by hype (SaaS being hot ≠ SaaS fitting you).
## Step 0 — Pick an audience with a "hair on fire" problem
Framework: Paul Graham "make something people want" + Amy Hoy Sales Safari.
Not a generic "target market" ("marketers"). Specific enough that you can answer:
- Where do they lurk? (which subreddit, which community, which YouTube channels they subscribe to)
- Which burning problem is still unsolved?
- Who are they paying today to solve it?
**Checkpoint before step 1:**
- [ ] You can answer: "my audience lurks in [3 specific places]"
- [ ] You can answer: "they pay $X/month to [A, B, C] to solve this problem"
Cannot answer → you are guessing. Stop and do step 1 more thoroughly.
## Step 1 — Sales Safari
Framework: Amy Hoy (stackingthebricks.com). Proof: Noko/Freckle Time, bootstrapped with Sales Safari.
**Do not survey. Do not interview. Go lurk.**
- Visit the 3 places the audience lurks (step 0)
- Read 100 posts/comments
- Find repeated phrases: "I wish X existed", "Why is X still so bad", "I've tried A, B, C — none work"
- Copy them verbatim into one doc — do not paraphrase, do not "translate" the pain
**Why not interview:** people lie when asked. People are honest when they complain among themselves.
**Checkpoint:**
- [ ] ≥20 verbatim quotes of the audience stating the pain
- [ ] ≥3 quotes REPEATING the same problem (from different people)
- [ ] The author of each quote (to reach out later)
**Adaptation by format:**
- Physical: Amazon 1-2 star reviews + niche subreddits
- Digital: Twitter/X, maker subreddits
- SaaS: G2/Capterra 1-2 star reviews + user subreddits
- Service: LinkedIn comments + professional Slack communities
- Course: Reddit "recommendations" threads + YouTube comment sections
- Community: existing free communities, look for "looking for people to do this with"
## Step 2 — Pick the product format
Match the problem to the format:
| If the repeated problem is… | Best format |
|---|---|
| "I have to do X over and over, it takes forever" | Digital template or SaaS |
| "I don't know how to do X" | Course or Service (coaching) |
| "I feel alone doing X" | Community |
| "I want X but there is no brand I trust" | Physical |
| "I need someone to do X for me" | Service |
| "I need an environment so I don't quit" | Community + Course combined |
**Hard trade-off:** high margin (digital/course) = high competition. Strong moat (community/physical brand) = slow to build. No free lunch.
## Step 3 — Cheapest prototype per format
Framework: Paul Graham "Do Things That Don't Scale" (paulgraham.com/ds.html). Cases: Airbnb photographing apartments, Stripe "Collison installation".
Prototype = a version that tests a hypothesis, NOT a version to sell at scale.
| Format | Cheapest prototype |
|---|---|
| Physical | Sample batch of 20-100 units, ship by hand, hand packaging. E.g. Emma Chamberlain launched D2C coffee bags before retail |
| Digital | Alpha Notion template shared with 10 friends |
| SaaS | Clickable Figma + Airtable + Zapier no-code (fake backend). Run the backend manually |
| Service | Do it yourself for 3 clients, log time in detail |
| Course | One live cohort, no recorded material, taught on Zoom |
| Community | Discord/Circle with 10-30 people for 30 days, each invited by hand |
**Cost cap: <$500 and <2 weeks for a prototype.** Over that = you are building, not prototyping.
## Step 4 — Pre-sell BEFORE building
Framework: Justin Jackson (justinjackson.ca). Cases: Marketing for Developers 5,000+ copies; Peak Design Everyday Messenger $6.57M on Kickstarter with 26,359 backers.
**You do not have a product yet. You are selling the idea.**
- A one-page landing describing the offer + CTA "pre-order $X" or "join the waitlist"
- A hard deadline (14-30 days)
- Send it to the audience (email list, social, community)
**Pass/fail thresholds:**
- ≥5% of the list joins the waitlist = weak signal (worth a prototype)
- ≥1% of the list pre-orders with money = strong signal (build now)
- <1% on both = kill the idea, back to step 1
**Never call a waitlist "validated".** Waitlist ≠ money. Only paid pre-orders count.
**Adaptation by format:**
- Physical: Kickstarter/Indiegogo pre-order
- Digital: Gumroad pre-sale, ThriveCart pre-order
- SaaS: AppSumo lifetime deal or "founding member" pricing
- Service: book a discovery call, take a deposit
- Course: cohort waitlist + 50% early-bird deposit
- Community: $99 lifetime founding member or first-100-only pricing
## Step 5 — Ship v0 to the first 10 people
Framework: Paul Graham "Do Things That Don't Scale".
- Ship to ≤10 people
- Onboard, support and deliver yourself
- No automation code, no hires, no big platform
- Log every conversation (any notes app / Loom)
**Checkpoint:**
- [ ] 10/10 have used it ≥1 time (not just "paid")
- [ ] ≥3 quotes about their experience (loved/hated)
- [ ] You know who dropped off and why
## Step 6 — Measure retention for 3 weeks, decide iterate or kill
Framework: Sean Ellis PMF Survey (pmfsurvey.com). Case: Superhuman 22% → 58% "very disappointed" (First Round Review, Rahul Vohra breakdown).
Send to the 10 v0 users:
> How would you feel if you could no longer use [product] tomorrow?
> (a) Very disappointed (b) Somewhat disappointed (c) Not disappointed
**Thresholds validated by Superhuman:**
- ≥40% "very disappointed" = PMF, scale
- 25-40% = iterate one round, retest
- <25% = kill, back to step 1
**Adaptation by format** (Ellis does not fit every format):
| Format | Retention metric instead of Ellis |
|---|---|
| SaaS / recurring digital / Community | Ellis "very disappointed" ≥40% |
| Physical one-off | Repeat purchase rate ≥30% within 30 days |
| Course | Completion rate ≥50% + NPS ≥50 |
| Service | Referrals — client brings ≥1 new client within 60 days |
| Digital one-time | Refund rate <5% + share rate ≥10% |
## Step 7 — Package + price with the Value Equation
Framework: Alex Hormozi, $100M Offers (acquisition.com/books). Case: Gym Launch $17M profit in year 1, $46M exit.
```
Value = (Dream Outcome × Perceived Likelihood)
─────────────────────────────────────────
(Time Delay × Effort & Sacrifice)
```
Packaging = raise the numerator, lower the denominator:
- **Dream Outcome**: reframe "12-week course" → "100 subscribers in 30 days"
- **Perceived Likelihood**: add case studies with concrete numbers, testimonials, a money-back guarantee
- **Time Delay**: shorten time to first value (a week-1 quick win)
- **Effort & Sacrifice**: bundle templates, cheat sheets, done-for-you assets
**Pricing anchor:**
- A single tier anchors badly. Use ≥2 tiers (entry + main) or 3 (entry + main + premium anchor)
- The premium tier is a psychological anchor, not a volume seller
- Ali's PTYA does it right: $1 / $995 / $4,995 — the $4,995 tier makes $995 look cheap
## Step 8 — Checklist before handing in
- [ ] A specific audience down to the subreddit/community
- [ ] ≥20 verbatim audience quotes
- [ ] A format that matches the problem (not a preference)
- [ ] Prototype cost <$500, <2 weeks
- [ ] Pre-sell with ≥1% paid pre-orders (not waitlist)
- [ ] v0 shipped to ≤10 people, by hand
- [ ] Retention test with a format-specific threshold
- [ ] Package with ≥2 tiers and a psychological anchor
- [ ] You re-read the framework and dropped ≥1 wrong assumption. Dropped none = did not read carefully.
## Forbidden
- Using "1000 True Fans" as an R&D playbook (it is north-star math)
- Quoting the Coolest Cooler ($13M) as a success case (bankrupt in 2019)
- Skipping Sales Safari when you are not an insider of the audience
- Building before pre-selling (unless it is <2 weeks of effort)
- Calling a waitlist "validated" (waitlist ≠ money)
- Using surveys/interviews instead of lurking (people lie when asked)
## Framework test — 3 real cases (showing it is universal)
### Case 1: Ali Abdaal PTYA Academy $995 (format: course)
| Step | What Ali did | Fit |
|---|---|---|
| 0 | Audience = doctors/knowledge workers tired of 9-to-5. Ali is the insider | ✓ |
| 1 | Skipped Sales Safari — lived the doctor→creator journey, knew the pain | ✓ |
| 2 | Format = course (90% margin, moat = doctor→creator brand) | ✓ |
| 3 | Prototype = cohort v1 in 2020 (~50 people, Zoom, not recorded) | ✓ |
| 4 | Pre-sell: waitlist + hard deadline, money before recording | ✓ |
| 5 | v0: taught cohort 1 live, moderated himself | ✓ |
| 6 | Iterate: cohort → v2 → v3 restructured into hybrid | ✓ |
| 7 | Package $1 / $995 / $4,995 — premium tier as anchor | ✓ |
**Framework fit 8/8.** PTYA is the template for the course format.
### Case 2: MrBeast Feastables (format: physical CPG)
| Step | What MrBeast did | Fit |
|---|---|---|
| 0 | Audience = Gen Z MrBeast viewers, hair-on-fire "I want to eat what Jimmy eats" | ✓ |
| 1 | Skipped Sales Safari — comment data from hundreds of millions of subs was enough | ✓ (edge case) |
| 2 | Format = physical CPG (moat = retail distribution in 30K+ stores) | ✓ |
| 3 | Prototype: 3 flavours launched D2C in 2022 before retail | ✓ |
| 4 | Pre-sell: launched in one video, 1M bars sold in 72h | ✓ |
| 5 | v0 D2C only before Walmart | ✓ |
| 6 | Iterate: reformulated in 2023, expanded SKUs (gummies, milk) | ✓ |
| 7 | Package: "ethically sourced" positioning + Golden Ticket Wonka promo | ✓ |
**Framework fit 8/8.** Applies to physical CPG for creators with a mega audience.
### Case 3: Nathan Barry / ConvertKit → Kit (format: SaaS)
| Step | What Nathan did | Fit |
|---|---|---|
| 0 | Audience = professional bloggers needing an email tool beyond Mailchimp | ✓ |
| 1 | Sales Safari: read blogs + Twitter where bloggers complained about Mailchimp | ✓ |
| 2 | Format = SaaS (moat = product depth + switching cost) | ✓ |
| 3 | Prototype: hand-coded PHP, run for the first <100 users | ✓ |
| 4 | Pre-sell: $79/month beta pre-orders, published "The Web App Challenge" openly | ✓ |
| 5 | v0: Nathan supported and onboarded each user himself | ✓ |
| 6 | Retention: nearly killed it in the first 6 months without PMF, iterated into the blogger niche | ✓ |
| 7 | Package: tiers by subscriber count, not by feature | ✓ |
**Framework fit 8/8.** Applies to bootstrapped SaaS. Nathan published the full journey at nathanbarry.com/15-lessons-15-million/ for verification.
## Numbers we refuse to print
This framework does not use these numbers because no primary source can be traced:
- "Justin Jackson MegaMaker pre-sale $50k" — no source
- "Peak Design Everyday Backpack $12M" — the real number is $6.57M for the Everyday Messenger (2016)
- "Slack/Dropbox passed the 40% test" — only Superhuman is confirmed (First Round Review)
- "altMBA lifetime revenue $123M" — an estimate, Wes Kao has not published it
- "acquisition.com portfolio $250M+ ARR" — self-reported by Hormozi, not audited
- "Ali PTYA $1.5M in the first 9 months" — no original post from Ali found; the verified number is "$1.9M largest cohort" from tryvirtually.com
## Note
If an output produced with this skill looks doubtful, re-read "Forbidden". Most creator R&D mistakes = skipping Sales Safari (step 1) or confusing a waitlist with pre-orders (step 4).
## LESSONS LOG
Read this whole section BEFORE writing the first word of output. Each row is a mistake
someone already pointed out; repeating it is worse than making it the first time.
An empty table means this skill has never been corrected.
**An empty table is honest — never invent sample rows.**
- **Spot feedback:** "that's wrong", "not like that", "don't…", "drop…", "next time…",
"from now on…", "I already told you", "still…", "how many times", or the user
hand-edits the output you just gave and sends it back.
- **What to log:** fix the output and sweep every place with the same problem — ALWAYS.
Logging is CONDITIONAL: only when the mistake is RULE-WORTHY, i.e. it would come back
with a different input. A ONE-OFF (wrong name, wrong number, wrong link, changed
preference) gets fixed, not logged. When logging, add EXACTLY ONE row to the table
below, in this very file, then say `Logged in creator-product-rnd LESSONS LOG: WRONG … → RIGHT …`.
The **Source** column is required and takes one of three values: `user` (the user
actually said it) · `review` (a review pass caught it) · `measured` (a tool run showed
it). Never present the agent's own conclusion as if the user said it — a made-up
source is worse than no row. The WRONG column describes observable behaviour.
- **Cap: 10 rows** — not an arbitrary number. The parent rule is `SKILL.md < 300 lines`;
this block takes 28 frame lines, plus one blank separator and 10 rows makes 39, so
the skill body must stop under 261 lines. Log full and a new row arrives → promote the
most repeated row to a named hard rule in the skill body and delete it from the log;
never delete a row just because it is old. If the two rules clash, `< 300 lines` WINS.
Check with `wc -l SKILL.md` (< 300) and `grep -c '^| 20' SKILL.md` (<= 10).
| Date | Source | WRONG | RIGHT |
|---|---|---|---|Free