Mid-sized enterprise application (e.g. internal order management, divisional specialist system): about 4–6 developers, 6–9 months, budget often from roughly EUR 150k—consolidating processes or replacing a legacy stack.
At this scale, a proof of value after 8–10 weeks is worthwhile. This is a working increment accepted by real users in a pilot unit. It includes KPIs such as lead time and defect rate, plus documented interfaces to neighbouring systems. This avoids modules that never reach production. We plan hypercare and key-user training alongside go-live.
Large platform (e.g. multi-tenant B2B marketplace, customer portal with many integrations): about 6–10 developers, 9–15 months, budget often from roughly EUR 300k. This needs explicit architecture, API strategy and staged approvals.
Here, integration windows and dependency chains are the bottleneck. Every external interface needs test data, a staging counterpart and a named vendor contact. We plan milestones around end-to-end journeys such as quote–order–payment–billing, not just component teams. Capacity targets are set early so performance testing does not start two weeks before launch.
Mission-critical system (e.g. real-time production control, core business platform): about 8–15 developers, 12–24 months, budget often from roughly EUR 500k. Availability, compliance and 24/7 operations dominate planning and QA.
For critical systems we define RTO and RPO and run repeatable disaster-recovery drills. We also set up clear incident escalation. Regulatory and internal audit requirements are met through traceable evidence: change logs, access records and approvals. Optional: certification or penetration test cycles so security is proven in practice, not just on paper.