.webp)
A first-time founder we spoke to last month had told his angel investor she would see a live MVP in 90 days. Sixty days in he had wireframes and a signed contract with an agency. The agency had promised delivery in "roughly 3 months" without a sprint plan. He was worried his 90-day promise would slip to six.
This article is written for him. What 90 days actually buys you in 2026. The realistic week-by-week sprint plan. The team you need. What has to be cut. Real cost bands. And the failure modes that turn a 90-day commitment into a six-month scramble.
If you promised a board or an investor a 90-day MVP, this is your survival guide.
What "90-Day MVP" Actually Means (and Does Not)
Ninety days is twelve weeks. Sixty working days. About the length of a company sprint quarter.
What 90 days buys you in 2026 with the right team and discipline: one clear user journey, 3 to 6 core screens or workflows, authentication and basic user management, payment integration (Stripe), essential admin functionality, and deployment to production with monitoring.
What 90 days does not buy you: complex role hierarchies, multi-language support, advanced analytics, in-app messaging, custom integrations with more than one or two external systems, native mobile apps, offline mode, or advanced customisation options.
The Y Combinator Library essays are unambiguous on this. Every experienced founder writes some version of "cut, cut, cut." That guidance is right. What you think is the MVP scope on day one is roughly 2x what your 90-day plan can actually ship.
The Week-by-Week Sprint Plan
Weeks 1-2: Discovery and locked spec. User stories written and prioritised. Wireframes drawn (hand or Figma). Tech stack decided. Team assembled. Spec locked at end of week 2. From this point, scope changes require explicit trade-offs.
Weeks 3-4: Design system, database schema, core auth. Design system built (colours, typography, component library). Database schema designed with future flexibility. User authentication implemented (Clerk or Auth.js). Basic landing page live.
Weeks 5-6: Core feature build (part 1). The single most important feature your MVP tests. If you cannot name it in one sentence, your scope is not tight enough. Build the primary user journey end to end. Rough edges acceptable.
Weeks 7-8: Core feature build (part 2) and admin. Complete the primary user journey with polish. Build the essential admin interface (user management, basic analytics). Data validation and error handling.
Weeks 9-10: UAT with 5-15 real users. Get real target users into the product. Watch them use it. Note everything that breaks or confuses. Fix critical issues. Not "polish" issues, critical issues. Small UI improvements only.
Weeks 11-12: Payment integration, deployment, launch. Stripe integration for subscription or transaction billing. Production deployment on Vercel or Fly.io. Monitoring and error tracking (Sentry). Marketing site refined. Launch to first paying customers.
This plan assumes the team never loses a week. Realistic 90-day sprints usually experience one or two disrupted weeks (illness, unexpected scope discovery, integration surprise). Build 2 to 3 days of buffer into weeks 5-8.
The Team You Actually Need
Three team configurations, with realistic outcomes at each.
Founder + one senior freelance developer. Realistic outcome: 5 to 6 months for meaningful MVP, or 90 days for very narrow scope (single-page tool, basic CRUD app). Not 90 days for real SaaS. Solo dev context switching costs kill the timeline.
Founder-as-PM + one senior developer + one designer. Realistic outcome: 90 days for tight-scope MVP (single core feature, 4-6 screens, simple auth, one integration). Founder must be genuinely available 3+ days per week.
Two developers + one designer + one PM + founder as product owner. Realistic outcome: 90 days for real MVP with meaningful core feature, admin, payments, and 5-15 user UAT cycle. This is the configuration that actually delivers.
The mistake founders make is choosing configuration 1 for budget reasons and then discovering month 4 that they cannot ship. Configuration 3 costs more upfront but almost always costs less in total programme spend by the end.
What Has to Be Cut to Fit 90 Days
Six categories founders reliably underestimate the cost of.
Complex user roles. Start with two roles (admin, user). Add roles later. Role complexity multiplies with every additional role.
Multi-language. English only for 90 days. Add languages post-launch based on user demand.
Multi-currency. One currency (usually USD or GBP). Multi-currency doubles the payment complexity.
Advanced analytics. Track a small set of essential events (signup, first action, subscription). PostHog or Mixpanel free tier. Custom dashboards later.
White-label capability. Not in the MVP. Ever. White-label is a v2 or v3 feature.
Multiple external integrations. One integration maximum in the MVP (usually Stripe). Second integration in weeks 11-12 only if truly essential. Additional integrations post-launch.
Real 2026 Cost Bands
Agency delivery. £45k to £85k over 90 days. Includes 2 developers, 1 designer, 1 PM, project management overhead. Best when founder has limited time to manage delivery. Predictable delivery. Higher total cost.
Founder-assembled freelance team. £25k to £55k. Founder recruits and manages 2 developers plus 1 designer. Cheaper than agency but requires founder to be available 3+ days per week for management, direction, and unblocking. See our related read on cost of building a marketplace platform for the two-sided variant.
Senior single freelance developer. £15k to £35k. Only viable for very narrow scope (single-page tool, basic CRUD SaaS). Not enough capacity for real SaaS in 90 days.
Recommended tech stack for 90-day sprint: Next.js on Vercel, Supabase or Neon for database, Stripe for payments, Clerk or Auth.js for authentication, Tailwind for styling, Sentry for error tracking. See Vercel Guides for reference architectures.
Add £2k to £5k per month for infrastructure, monitoring, and support tools during and after the sprint.
What We Learned Building IELTSArena's First 90 Days
IELTSArena is our AI IELTS preparation platform. The first 90 days shipped writing feedback with human-comparable band scoring. Two lessons.
Ruthless scope discipline is the enabling constraint. The initial scope had 40+ features. We shipped 6 in the first 90 days. The other 34 arrived over the following 18 months. Founders who insist on scope discipline early ship on time. Founders who add scope in week 6 do not.
Golden test set from day one. We started building the evaluation harness in week 3, not month 3. Every feature we shipped had a set of correct-outcome examples we could re-run when we made changes. This is what let us move fast without breaking things.
You can see IELTSArena at our portfolio. If you want a candid conversation about a 90-day scope for your idea, get a 90-day MVP scope from WhiteStone.
Common Failure Modes
Three failure modes we see repeatedly.
Scope creep in week 5. Founder sees demo, wants "one more feature." Two of those requests kill the 90-day timeline. Say no.
Choosing solo developer for budget reasons. Cheap upfront, slow in reality. Six-month delivery costs more than 90-day delivery in most cases.
Skipping UAT weeks 9-10. Team wants to keep building. Founder does not push back. MVP launches on day 90 with issues that would have been found by 10 real users. Retention suffers.
Frequently Asked Questions
Can a real SaaS MVP actually be built in 90 days?
Yes with a tight scope, disciplined execution, and the right team (2 developers + 1 designer + 1 PM). The MVP will have one core feature, 4-6 screens, basic auth, one payment integration, and essential admin. It will not have complex roles, multiple integrations, multi-language, or advanced analytics.
What has to be cut to fit in 90 days?
Complex user roles (start with two), multi-language, multi-currency, advanced analytics, white-label capability, and integrations beyond one or two. Also cut in-app messaging, offline mode, and any nice-to-have features. Founders who cannot cut ship in 180 days.
How many developers do I need for a 90-day MVP?
Two developers plus one designer plus one PM. Solo developer + founder-as-PM realistically ships in 5-6 months, not 90 days. Single senior freelance developer works only for very narrow scope.
What is the realistic cost of a 90-day SaaS MVP?
Agency: £45k to £85k. Founder-assembled freelance team: £25k to £55k. Senior single freelance developer (narrow scope only): £15k to £35k. Add £2k to £5k monthly for infrastructure and tooling.
What tech stack works for a 90-day sprint?
Next.js on Vercel, Supabase or Neon for database, Stripe for payments, Clerk or Auth.js for auth, Tailwind for styling, Sentry for error tracking. Modern stack, well-documented, fast to build with. Vercel Guides has reference architectures.
The One Thing to Remember
Ninety days is not a lot of time. What ships in 90 days is one core hypothesis, executed cleanly, with real users testing it in weeks 9-10 before launch. What does not ship is your full product vision. Founders who accept this ship on time. Founders who try to fit v1 of the dream into 90 days ship v0.5 in 6 months and disappoint everyone.
If you want a candid conversation about a 90-day scope for your idea, browse our MVP development services or come to the scoping call.
