1 What features does your project need?
Check only what v1 genuinely requires — not nice-to-haves. Each adds real hours.
2 What's your tech stack?
Pick the closest match to what you're planning to build with.
3 How well do you know this stack?
Honest answer. This is the single biggest variable in any time estimate.
Select features above to see your estimate
← or load a saved scope from a shared linkWhy most project estimates are wrong
Developers routinely underestimate side projects by 2–4x. The culprit isn't the fun parts — it's the invisible infrastructure. Authentication, deployment, error handling, responsive design, and "just one more feature" syndrome add up fast.
The three features that hurt most: user auth (even with Clerk or Auth0, it's 8–16 hours to get right), payment processing (Stripe seems simple but webhooks, edge cases, and compliance add 12–24 hours), and real-time features (WebSockets introduce state complexity that ripples through your whole app). These are where most "2-week projects" become 3-month projects.
The best scope for a weekend project: one clear user action, no auth (or third-party auth), data in localStorage or a simple free database, deployed URL by Sunday. Everything else is scope creep. The version you're embarrassed to ship is usually better than the version you never ship.
Scope and estimation questions
How long does it take to build a side project?
It depends on features. A static page or simple tool takes 2–4 hours. A CRUD app with auth takes 16–30 hours. A full SaaS with payments, multi-user support, and an admin dashboard takes 80–200+ hours. The biggest factor is how many "infrastructure" features you need — auth, payments, and real-time are each 10–20 hour investments on their own.
What makes a good weekend project?
One clear user action, no login required (or handled by a third-party service in 1 hour), data in localStorage or a simple free database, deployed URL by Sunday evening. Scope it to 8–16 hours. If you need more than 3 features to describe your core value prop, you're building a week-long project.
Should I add authentication to my side project?
Only if users need to own their data across devices or share it with others. Auth adds 8–16 hours even with a service like Clerk or Auth0. Many successful tools work without login — data lives in localStorage or is shared via URL. Ask yourself: would this be useful without accounts? If yes, skip auth for v1 and add it when real users ask for it.
How do I estimate software project time correctly?
List every feature including the boring ones. The "infrastructure" work (auth, deployment, error handling, responsive design) takes as long as the interesting parts. Add up estimates per feature, multiply by your experience level (beginners: 2–3x), then add 20% for debugging. Most developers underestimate because they only count the interesting work and forget the invisible bits.
What should I cut to ship my side project faster?
Highest-ROI cuts: (1) Skip auth — use localStorage or URL-based access. (2) Skip payments — launch free first, add monetization when someone asks to pay. (3) Skip the admin dashboard — use your database's built-in UI (Supabase, PlanetScale). (4) Make it polling instead of real-time — refreshing every 30 seconds is fine for v1. (5) Skip tests — write them after the core works. Each saves 4–20 hours.
How long does it take to build an MVP?
A real MVP — the smallest thing that proves the core value — typically takes 20–60 hours for a solo developer. Most so-called MVPs are overbuilt: they include auth, dashboards, and settings nobody asked for. The test: can you describe what users DO in one sentence? If you can't, you're building a product, not an MVP.
Does experience really change the estimate that much?
Yes — a developer learning a new stack should expect 2–3x more hours than an experienced developer for the same features. The bottleneck shifts from "what to build" to "why is this auth callback not firing." That's not a flaw — it's how learning works. Which means beginners should aim for even smaller scopes: finish one 4-hour project, then build the next one. Finishing compounds; ambition that never ships doesn't.