Skip to main content
dewgo

MVP

Spark MVP. What's actually in the $3,000

The Spark MVP tier is $3,000 flat. People ask what fits in that budget. The honest answer is "more than you think, less than you wanted." Here is the exact scope. What is in. What is out. Why those lines are where they are.

5 min read · published May 21, 2026

What is in scope

A real working web product. Not a clickable mockup, not a no-code prototype. A real Next.js app on Vercel that users can sign up for and use.

  • One core feature. The single thing your product does, working end to end.
  • Auth. Email and password, magic link, or Google OAuth. Pick one.
  • A database. Supabase, with row-level security set up correctly.
  • A landing page that explains the product and converts to sign-up.
  • Hosted on a domain you own. SSL. Real email sender via Resend.
  • Analytics so you can see what users do.
  • Stripe payments if your business model needs them. Subscription or one-time.

What is not in scope

Anything that is not the one core feature.

  • Native iOS or Android apps.
  • Multi-language support.
  • Complex admin dashboards. We give you Supabase Studio, which is enough for v1.
  • A second feature. We will write a v2 plan, but we will not build two features in two weeks.
  • Integrations beyond the obvious ones (email, Stripe). Custom CRM connectors are a separate scope.
  • Pixel-perfect bespoke branding. The design uses our component library plus your colors and logo.

The four-week timeline

Spark is the one project type we extend beyond two weeks when the scope demands it. Most MVPs take three to four weeks: one for the core feature, one for auth and data, one for polish and launch, and a fourth that is half buffer and half QA.

  • Week 1. Architecture, auth, database. The non-negotiable plumbing.
  • Week 2. The core feature. End to end, with real data.
  • Week 3. Landing page, payments, polish, analytics.
  • Week 4. QA, fixes, launch, the runbook handoff.

What we use

Next.js 15 with the App Router. Tailwind. Supabase for auth and database. Resend for email. Stripe if you need payments. Vercel for hosting. Vercel Analytics or Plausible for usage data. The same stack we use on dewgo.org.

We do not use Webflow, Bubble, or any no-code platform for Spark. Custom code is faster to scale, easier to hand off, and not lock-you-in.

What you get at the end

A live product with real users you can sign up. The full source code in a GitHub repo you own. A deployed Vercel project under your account. A Notion runbook with admin links, env var setup, and the answers to the eight questions every founder asks in week three.

Plus a written v2 scope. After two weeks of real users, we know what is broken, what works, and what to build next. The v2 scope is your starting point if you want to keep going.

When $3,000 is the wrong number

When you need a mobile app from day one. When the product is a marketplace with two sides that both need a real experience. When you need integrations with three or more SaaS tools. Those projects start at $8,000-$20,000 and run six to ten weeks. Different conversation.

Keep reading