Every founder asks the same question on the first call: how much will this cost? Every honest answer starts the same way: it depends. Then somebody quotes a confident-sounding figure anyway, the founder believes them, and twelve months later there's a gap between the budget and reality wide enough to fall through.
This is the version of that conversation we give founders on our own calls: how we price a first build, how bigger platforms get scoped, the levers that move the number, the hidden costs nobody quotes, and a way to budget the 18 months in front of you instead of just the build.
How SaaS builds are priced in 2026
Forget "small / medium / large" — those map to nothing. A more useful way to think about it is in stages, where each stage is scoped and priced on its own.
Stage 1 · A focused first version — typically USD 3k–15k+
One core workflow, one or two user roles, one happy path. Auth, simple billing (a single Stripe plan), a dashboard, and the feature that's the actual reason the product exists. This is the range our typical initial projects fall into: a smaller, sharply scoped v1 sits near the bottom, while more roles, integrations or AI features push it toward the top and beyond.
At this stage you're testing whether the problem is real and whether people will pay. You will rebuild parts of it later. Plan for that. The v1s that never get rebuilt are usually the ones that didn't find traction — and that's the cheapest possible way to find out.
Stage 2 · Production SaaS — built in fixed-price milestones
Multi-tenancy, plans and seats, role-based access, an audit log, a customer admin console, observability, analytics, and the feature breadth a paying customer expects. Rather than one large quote, we scope this as a series of milestones, each with its own fixed price and a working release at the end.
The total depends on how many milestones you actually need, which is exactly why it's split up: you can stop, reprioritise or change direction between milestones based on what customers tell you, instead of paying up front for features nobody has asked for yet.
Stage 3 · Platform / enterprise — scoped individually
SSO (SAML / OIDC), org management, granular permissions, SCIM provisioning, usage-based billing, a public API, SDKs, partner integrations, regional data residency, and SOC 2 readiness baked into the architecture. Work at this level starts with a paid discovery phase that produces exact requirements, architecture, milestones and a fixed cost in writing, then is delivered in phases.
At this stage the costs that matter most aren't engineering hours — they're the design choices that decide whether the system stays maintainable as you grow or rots into a rewrite.
The seven things that actually drive cost
Most quotes are sensitive to the wrong variables. Here's what actually moves the number, in roughly the order it does:
1. Number of distinct user roles
Every role is its own product surface. One role is an MVP. Three roles is a real product. Six roles is an enterprise platform with real RBAC and probably an admin role on top of that. Each one adds screens, permissions, edge cases, and tests.
2. Whether you need multi-tenancy from day one
Single-tenant feels cheaper for the first few customers and bites you later. Building multi-tenant from day one costs a bit more up front but avoids a painful migration once you have real customers and real data. For almost every B2B SaaS, start multi-tenant.
3. Compliance scope
SOC 2 readiness, HIPAA and PCI each add work and change how the system is built, so they need to be in the scope from the start rather than bolted on. GDPR data export and deletion is now baseline, not a feature.
4. Real-time vs request-response
WebSockets, presence, live cursors, collaborative editing — all of these cost noticeably more than the equivalent request-response feature. Don't ship them in v1 unless they're the product.
5. AI features that aren't demos
A GPT wrapper is a weekend. A production AI feature with eval suites, observability, fallback logic, prompt versioning, and cost ceilings is a project of its own on top of the rest of the product.
6. Mobile + web
Mobile from day one adds a second product to build, test and release. If you can ship a great responsive web app first and add mobile once you have traction, do that. Most B2B SaaS doesn't need mobile in v1.
7. The integrations list
Each non-trivial integration (Salesforce, HubSpot, QuickBooks, Stripe Connect, an industry-specific API) is real engineering work with its own edge cases. The list of integrations on a website rarely matches the integrations actually live; the gap is the integration backlog.
The hidden costs nobody quotes you
These line items will surprise you if nobody warned you about them up front. Ask any vendor whether each one is included in their quote.
- Design system + component library — without one, every new screen takes longer to build and looks less consistent.
- Production infrastructure — staging, preview environments, CI/CD, secrets management, backups. Some setup work plus ongoing cloud spend.
- Observability tooling — Sentry, Datadog, PostHog or LogRocket. Usually a monthly subscription that grows with usage.
- Customer email infrastructure — transactional email (Postmark / Resend), domain authentication (DKIM, DMARC) and inbox deliverability.
- Legal — Terms of Service, Privacy Policy, DPA, an MSA template, and an Impressum if you sell into DACH.
- Branding & marketing site — logo, brand guidelines, marketing site, social cards.
- Onboarding & support tooling — Crisp, Intercom, help docs, in-app guides, and someone's time to run them.
- Year-1 cloud spend — modest at low traffic, but it grows quickly if you store files or run AI inference. Budget for it monthly from day one.
Build vs hire vs partner — the honest comparison
We're a development partner, so take this with the appropriate grain of salt. Check local salary data on sites like Glassdoor and Levels.fyi for your own market.
- Hire in-house: salaries, benefits, equipment and management overhead, plus months of recruiting before anyone writes code. Worth it once the product is your long-term core and the roadmap is stable.
- Development partner: no recruitment lag and hand-over when you're ready. With us, a typical initial project is USD 3k–15k+, and larger scopes are delivered in fixed-price milestones — no hourly billing.
- Lowest-bid hourly shops: the headline rate looks cheap, but open-ended hours plus rework can make the real total far higher than the quote. If you go this route, insist on fixed milestones and code in your own repository.
- Build it yourself: if you're a technical founder with months of runway and no employer, your time has an opportunity cost that's invisible but very real. Budget that into the comparison.
Budget for the 18-month total cost — not just the build
The number that matters isn't the build cost — it's the total cost of getting to product-market fit and your first paying customers. A rough framework:
- Build (months 0–4): your stage 1 cost from above.
- Iteration (months 4–9): 30–50% of build cost again — fixing what users actually complained about.
- Operations (months 0–18): 15–30% of build cost per year — observability, maintenance, security patching.
- Growth feature gap (months 9–18): 50–80% of build cost again — the features paying customers ask for that aren't in v1.
Using that framework, a USD 10,000 first build works out to roughly USD 20–28k over 18 months, before third-party tools and cloud spend. If that number scares you, halve the v1 scope before you halve the budget.
What we'd ask you before quoting
If a vendor quotes a number before they've asked these questions, the number is invented:
- How many distinct user roles, including admin?
- Multi-tenant from day one, or single-tenant?
- What does success at month 6 look like — feature list or revenue?
- Compliance posture — SOC 2, GDPR, HIPAA, none?
- Real-time features — collaboration, presence, notifications?
- AI features — and is the eval / observability scope included?
- What integrations are non-negotiable in v1?
- What's the in-house team size on your side, today and in 12 months?
The TL;DR
A focused first version of a SaaS product typically falls in the USD 3k–15k+ range with us. Anything larger — full production SaaS or an enterprise platform — should be scoped in fixed-price milestones after a proper discovery, not quoted as one big number from a one-line brief.
Budget for what you'll actually need over 18 months, not just what's on the first quote. Halve the v1 scope before you halve the budget. And put the comparison with hiring in-house honestly on the table before you decide.