EntrepreneurBible
All hubs

Product

MVP — what to leave out

The point of an MVP is not to impress — it's to learn what you'd otherwise spend months guessing about. Most founders over-build their first version because they're scared of looking unprofessional. The customers who matter want a thing that solves their problem; the rest can wait. This hub walks through the three MVP shapes, how to scope by what to leave out, and how to ship in weeks instead of quarters.

Last updated May 21, 2026

Who this is for

Founders who've validated a problem and need to ship a first version that learns, fast, without burning the runway.

What you'll learn

  • Three MVP shapes — concierge, Wizard-of-Oz, feature MVP — and which one fits you
  • How to scope an MVP by what to leave out
  • Realistic timeline: 2-6 weeks, not 6 months
  • What to measure in the first 30 days post-launch
  • When the MVP is the right architecture, and when to rewrite
Get the MVP plan template

Three MVP shapes — pick one

Concierge MVP — you do the work manually for the first 10-20 customers. No software. Customer doesn't know (or cares). You learn the exact workflow, pricing, and edge cases before writing code. Perfect for service-shaped problems.

Wizard of Oz MVP — software UI on the front; humans (often the founder) do the work behind. The customer experiences "the product"; you experience "an unscalable thing that's teaching me everything." Used by Stripe, Airbnb, DoorDash in their first 6 months.

Feature MVP — actual software, narrowest possible feature set. Right when the workflow is clear (you've already done concierge or Wizard of Oz), or when the value is in a specific software capability that humans can't fake.

Most founders default to Feature MVP. Most successful founders start with one of the first two.

Scope by leaving out

Make a list of everything you want to build. Cross out anything that isn't needed for the customer to:

  1. Sign up and start using it
  2. Complete one core action
  3. Pay you (if you're charging)

What's left is the MVP.

What to leave out (almost always): admin dashboards, settings pages, "advanced" features, integrations beyond the one most important, mobile apps if web works, internationalisation, password reset (use magic links), avatars, themes, multi-user, billing UI (use Stripe Checkout), notifications beyond one transactional email.

Don't optimise for: scaling to 10,000 users, code aesthetics, the right framework choice, multi-tenancy, complex permissions. You'll throw 80% of the MVP away anyway.

A useful rule: if you can't draw the MVP's full UI on one piece of paper, scope is too big.

2-6 weeks, not 6 months

Realistic MVP timelines:

  • Concierge / Wizard of Oz: 1-2 weeks (mostly setting up Stripe + a landing page)
  • Feature MVP, founder-built: 4-6 weeks (if focused, no second job)
  • Feature MVP, contracted out: 8-12 weeks (and harder to iterate)

If your MVP is taking 3+ months, scope is wrong. Cut a feature; ship; learn; add it back if needed.

The 6-week MVP plan:

  • Week 1: paper prototype + 5 customer reviews (no code yet)
  • Week 2-3: core flow (sign up → main action → done)
  • Week 4: payments (Stripe Checkout, no custom UI)
  • Week 5: ship to first 5 customers
  • Week 6: fix the things that immediately broke

Anything else (the second feature, the second persona, the second segment) waits.

Step-by-step action plan

Do these, in order

  1. 1Decide which MVP shape (concierge / Wizard of Oz / feature) fits your problem
  2. 2Draw the MVP's full UI on one piece of paper — if it doesn't fit, cut scope
  3. 3Set a 2-6 week deadline; nothing scope-creeps in until v2
  4. 4Ship to 5 customers, watching them use it via Loom or live call
  5. 5Decide one metric to track for the first 30 days (activation, not vanity counts)

Frequently asked questions

What stack should I use for an MVP?
The one you (or your co-founder) are fastest in. Next.js + Supabase or Rails + Postgres are common defaults but the right answer is whichever you can ship in 2 weeks not 8.
Should the MVP be a no-code build?
If the founder is non-technical: yes. Bubble, Glide, Softr, Webflow + Airtable cover most early MVPs. Move to code only when no-code constraints start dictating the product.
How much should an MVP cost?
Founder-built: $0-2k (Vercel + Supabase + Stripe). Contracted: $5-25k for a focused feature MVP. If you're being quoted $50k+ for a first version, scope is wrong (or vendor is).
When do I rewrite the MVP?
Three signals: (1) every new feature breaks two old ones, (2) onboarding a new dev takes more than a week, (3) you can articulate what the product really is and what the MVP got wrong. Until then, ship more features on the existing base.

Related resources

Idea-to-MVP Sprint Map

Related tools

Business Idea Scorecard

Related templates

MVP Plan

Related courses

Product Velocity for Founders

Related hubs

Business Idea Validation

A direct guide to validating a business idea before you build. Interviews, smoke tests, concierge MVPs, pre-orders, and the evidence hierarchy.

Customer Discovery

Run customer discovery interviews that produce signal, not validation theatre. The 5-question framework, how many to run, and how to spot fake demand.

Startup Launch

How to launch a startup without faceplanting. The soft-launch first principle, when Product Hunt actually helps, the 14-day pre-launch checklist, and what to measure in week 1.

Product-Market Fit

What product-market fit actually feels like, the four signals that prove it, the 40% test, and what changes after — without the cargo-cult mythology.