Video summary

I Built 2 SaaS Apps to $50K MRR: Here's My 90-Day Playbook

Main summary

Key takeaways

Business

Core business thesis / playbook

  • Start with the customer, not the product: validate demand by gauging how urgently customers want the solution and whether they’ll pay.
  • Sell before you build (or at least de-risk with paid pilots), using high-intent lead generation and clear offers.
  • Use proven market “cloning” selectively: find profitable competitors in the target geography/category, then adapt only what’s necessary.

Fail → Learn → Win (Coco AI and Goji Berry)

1) Failed attempt: “build first” + wrong customer readiness

They tried to expand a profitable concept (American “text-to-buy” SMS) into France/Europe by:

  • Rebuilding to WhatsApp (since SMS buying wasn’t a thing in Europe)
  • Launching with only 1–2 pilot customers

Why it failed (customer validation gap):

  • Customers weren’t ready to pay via text/WhatsApp purchase flow
  • Product-market fit was missing because they didn’t talk to customers in advance

Outcome:

  • Spent 6 months without traction.

2) Breakthrough attempt: clone proven Shopify apps + ship fast

Strategy shift:

  • Stop copying a whole unrelated approach; instead clone profitable Shopify apps already working in Europe

Execution:

  • Identified Shopify app winners making ~$2M–$3M/year
  • Cloned a competitor app and launched in ~2 weeks
  • Implemented two main feature changes

Early traction:

  • Reached out to contacts; within ~2 days, got 5–10 clients
  • The offer fit a market where few/no competitors were selling this in France

Team / operating model (division of labor)

  • Role allocation emphasized:
    • Marketing-led execution
    • Product & code support via co-founders/dev

Coco AI team size and roles:

  • Initially: 4 people — ~2 in marketing, 1 product, 1 coding
  • Later: 3 people — 1 marketing, 1 product, 1 coding

Key KPIs & targets mentioned (with timeline anchors)

Coco AI

  • Grew to $50K MRR in ~6 months
  • Cold outreach-driven; close rate on demos stated around 60–70%
  • ICP: Shopify stores making > $10K/month in Europe
  • Later scaling/selling:
    • Sold for seven figures about ~2 years later
    • Outreach started in 2023; sold by 2025
  • Scaling friction:
    • Demo/close rates dropped to ~15–20% in non-French markets (e.g., Germany/Spain)
  • CAC/LTV/churn: no numeric values provided (only qualitative “low churn” mentioned)

Goji Berry

  • V1 (UGC/AIGC concept): hit up to ~$500 MRR in ~2 weeks, then shut down due to lack of interest/evolution path
  • V2 (notetaker style, “Fireflies-like”):
    • Built MVP in ~2 weeks
    • Sales were “awful”
  • V3:
    • Sold in Feb, launched in Sep (~6 months total), then scaled:
    • 0 → $33K MRR in ~4 months

Distribution / traction metrics

  • Reddit: ~11 million impressions in ~4 months

Cashflow / sales de-risking

  • Goji Berry sold $7,000/month (or “$7K MRR”) before coding first feature (sell-first strategy)

Frameworks / processes / playbooks extracted

Customer-first validation

  • Ask:
    • How much do they want it?
    • How much will they pay?
  • Don’t guess readiness—talk to customers early.

Copy-to-learn (but validate real unfair advantages)

  • Find profitable apps already selling
  • Use market signals:
    • app store data, download/store intelligence, founder posts
  • Clone quickly (target: 2-week launch mentioned)

Sell-before-build

  • Build an offer around what customers already want (high-intent leads)
  • Charge for value first; then automate fulfillment

High-intent outbound as the core machine

  • Target audiences already “in-market” (e.g., commenters on WhatsApp/email marketing posts)
  • Prefer “yes” actions (blueprints/trials) over intrusive “demo” asks

Distribution flywheel (add channels gradually)

  • Start with one channel heavily (e.g., Reddit)
  • After it works, add more:
    • X, SEO, YouTube, Trust (ads marketplace), later Threads

Cold outreach KPI loop

  • Primary growth lever: volume + offer conversion
  • Improve demo-close rate via messaging/ICP matching

Concrete GTM tactics & actionable recommendations

1) ICP definition (made it measurable)

  • ICP: Shopify stores > $10K/month revenue in Europe
  • Bonus segment: stores using AI tools (higher likelihood to act)

2) Outreach volume operations

  • Daily outbound stated:
    • ~100 calls/day
    • ~1,000–2,000 cold emails/day
    • Plus LinkedIn outreach
  • Demo conversion assumption:
    • When someone asks for/gets a demo: close rate ~60–70%
  • Lead sourcing tools/channels mentioned:
    • LinkedIn + Sales Navigator
    • Store Leads
    • Fox Leads
  • Lead list building based on:
    • store size, AI tool usage, and category

3) “Email customer service instead of founder” workaround

Instead of emailing founders:

  • Email customer service with a demo video and ask them to forward to the CEO

Claimed benefits:

  • Nearly always replies
  • Avoids spam issues
  • Increases meeting bookings via internal routing

4) Replace “Ask for a demo” with “Offer a blueprint”

Offer structure:

  • Email looks demo-adjacent but ends with:
    • “If you agree, we’ll send you a blueprint”

Blueprint details:

  • ~10 pages of actionable value
  • Booking link at the end, but many convert via free trial instead of a call

Why it works:

  • Easier “yes” than scheduling a demo; higher conversion to the next step

5) Sell-first mechanism for Goji Berry (lead productization)

Reframe the problem:

  • Not “AI tool for AI’s sake,” but high-intent lead generation as the deliverable

Offer packaging:

  • Charge $200–$300/month for 100–200 high-intent leads

Lead generation method (manual → automated):

  • Start manually:
    • search LinkedIn queries by niche (example: “dental clinic nurse”)
    • extract interacting/qualified prospects
    • send via email/phone list
  • Later:
    • hire via Fiverr with an exact blueprint for production

6) Distribution stacking (grow without betting everything on one channel)

  • Heavy posting in Reddit first (stories + proof + link)
  • Then add:
    • X, SEO, YouTube, Trust; later Threads

Reddit content template:

  • Story of what happened (including outcomes)
  • Proof (screenshots/rejection/intermediate evidence)
  • Link to product

“Acceleration hack”:

  • Post support via a friend group for quick upvotes/shared reposting

7) Influencer + newsletter sponsorship guidance (high level)

  • Mostly tested French influencers/newsletters for easier access via networks
  • Newsletter results:
    • Profitable within ~2–3 months
    • Attribution to long-term conversions was difficult
  • Influencer results:
    • LinkedIn influencers were stated as more profitable than newsletters
    • Pricing advantage: easier negotiation; better economics than newsletters
  • Trust management:
    • Used paid visibility (Trust platform / X ad)
    • Claimed profitability “every two or three days” via founder retweets

8) Affiliate program lever

  • Commission:
    • 30% lifetime
    • planned increases to 40% → 50% → 60%
  • Rationale:
    • higher commission → more affiliate effort → more traffic, even from non-affiliate buyers
  • Current contribution:
    • affiliates ~10–15% of revenue (not yet a primary focus)

Why the business was sold (strategic reasoning)

  • Market saturation in France:
    • reached out to nearly everyone in ICP
  • Geographic scaling difficulty:
    • demo close rate dropped to ~15–20% outside France (e.g., Germany/Spain)
    • harder to secure demos and close them
  • Team motivations + lack of future vision:
    • two team members wanted to sell
    • founder lacked a clear expansion vision beyond the initial steps
  • Deal process:
    • marketplaces (Acquire/Flippa) offered low multiples without 1-year revenue history
    • fast sale (~2 months) from directly reaching out to successful multi-app Shopify founders
    • also from a LinkedIn buyer query he noticed

Biggest mistakes (lessons)

  • Building something customers didn’t want
  • For Coco AI:
    • product mistakes were “tons”
    • major operational risk: influencer spending on fake influences (lost money; described as “cost of doing business”)
  • For Goji:
    • V1/V2 were killed after learning sales/interest were weak
    • V3 corrected by anchoring on a proven high-intent offer

90-day “start over” plan (step-by-step, as stated)

Goal: reach deal/traction by validating demand quickly, then scale distribution.

  1. Identify your unfair advantage (skills + what you can do better)
  2. Study the market: who is winning, who’s raising vs bootstrapping
  3. Build a simple offer doc (PowerPoint / 7 slides):
    • pain points, solution, how it works, price
  4. Pre-sell via discovery:
    • message 100–500 people
    • get them on calls; capture real pain points (often differs from your slide assumptions)
  5. Get paying demand:
    • convert “wants to pay” into service delivery first or plan the build
    • bring co-founder/dev if needed
    • validate with 5–10 people
  6. Launch GTM with outbound first:
    • cold email + LinkedIn outreach
    • budget ~$300–$500 to run tools at minimum
    • aim for ~10 more demos
  7. Add inbound once outbound works:
    • post on LinkedIn, X, Reddit (seek feedback/reviews)
  8. Month structure:
    • Month 1: outreach focus
    • Month 2: outreach + inbound
    • Month 3: add additional traffic sources (paid/influencers), depending on cash available

Presenters / sources

  • Roman (SaaS founder; primary interviewee)
  • Host / Interviewer (name not provided in subtitles; questions asked throughout)

Original video