A web application estimate starts with behavior.
A marketing website publishes information. A web application lets a customer, partner or employee perform a task, manage data or return to a recurring workflow. That behavior creates states, permissions, validation, recovery, notifications and operational responsibility that page counts do not describe.
Before estimating, define the primary user, the repeated job, the business outcome and the smallest end-to-end release that can produce evidence. A feature list without those boundaries usually becomes a budget for uncertainty.
Indicative investment bands for planning.
These are broad USD planning bands, not a fixed North Growth Lab quote or a universal market price. Existing infrastructure, security requirements, third-party services, product uncertainty and the delivery model can move a project outside these ranges.
| Scope | Typical fit | Indicative investment |
|---|---|---|
| Prototype or narrow internal tool | One workflow, limited roles and controlled users | $10,000–$30,000 |
| Production MVP | Customer portal or SaaS workflow with accounts and integrations | $30,000–$100,000 |
| Complex platform | Multiple roles, workflows, systems, compliance or mobile products | $100,000+ |
The largest cost drivers are structural.
- Number of distinct user roles and permission boundaries.
- Complexity of the core workflow, exceptions and recovery states.
- Existing data quality, migration and synchronization requirements.
- External APIs, payments, identity providers, CRM or operational integrations.
- Administration, reporting, audit history and customer-support tools.
- Security, privacy, accessibility and regulated-industry requirements.
- Native iOS or Android applications sharing the same backend.
Define the MVP as a complete loop, not a smaller feature pile.
An MVP still needs secure access, understandable errors, usable administration and measurement. Removing those foundations may reduce the demo cost while making the product impossible to operate or evaluate.
| MVP question | Useful answer |
|---|---|
| Who uses it first? | One named segment with a recurring problem |
| What do they complete? | One end-to-end job with a clear success state |
| Why do they return? | Stored value, collaboration, status or repeated need |
| What does the business learn? | Adoption, completion, conversion or retention evidence |
| Who operates it? | Named owners for support, data and release decisions |
Compare proposals by responsibility, not hourly rate alone.
- Is product discovery and workflow definition included?
- Who owns UX, content, data model, infrastructure and quality assurance?
- Which integrations and migration responsibilities are explicit?
- How are security, privacy and accessibility requirements accepted?
- What analytics and operational tools exist at launch?
- Who owns the code, accounts, documentation and deployment process?
- What support and iteration are available after the first release?
Control cost by reducing uncertainty early.
A short discovery and prototype phase can expose permission conflicts, missing states, weak third-party APIs and operational assumptions before they are embedded in production code. This is not paperwork before development; it is the cheapest place to change the product.
The responsible estimate is the smallest scope that can be launched, operated and measured. Once real users complete the workflow, the roadmap can expand from evidence instead of internal prediction.