When to build
Custom software is an operational decision, not a cause
Ready-made tools solve many problems better and for less. Building starts to make sense when their limits create meaningful cost, risk or rework for an established operation.
The problem
The problem is rarely just the subscription price
Fragmented processes
When data and responsibilities are spread across tools, reconciliation, rework and poor visibility may cost more than the subscriptions.
A process shaped around the tool
The team creates parallel controls because the software does not support important details of how the work is done.
Dependency with no path forward
Limited exports, fragile integrations or a legacy system without an owner make every change slow and risky.
The comparison
Compare total cost, not an invoice with a promise
Ready-made tools have subscription, rollout, integration and operational costs. Custom software has discovery, build, migration, infrastructure and maintenance costs.
A decision is only defensible when those costs, the risks and the value of the problem enter the same calculation. There is no universal payback period.
What belongs in the calculation
- Subscriptions, rollout and price increasesTools
- Integrations and remaining manual workOperation
- Build, migration and infrastructureSystem
- Maintenance, evolution and riskTotal cost
The figures depend on the process, the comparison period and operational risk. Sothen examines the assumptions before recommending a build.
Cost of the current process
Include manual hours, errors, waiting, rework and decisions made without reliable information.
Cost of building and maintaining
Consider discovery, development, migration, cloud services, external suppliers, security and continuous improvement.
Cost and risk of transition
A change requires team participation, data validation, training and a plan to keep the operation running.
Value of control
Consider the value of setting priorities, connecting data and changing technical teams without losing the system.
Decision criteria
When replacing makes sense, and when it does not
It makes sense when
- Critical processes depend on spreadsheets, manual tasks or parallel controls
- Important data is spread across systems that do not communicate
- The current system prevents necessary changes to the operation
- The process is relatively stable and someone is responsible for it
- The cost of the problem is significant enough to justify the investment and transition
It does not make sense when
- An established tool already solves the problem well
- The process still changes completely every week
- A small automation or integration can solve the bottleneck
- Nobody on your side is responsible for the operation
- The main motivation is simply to stop paying subscription fees
Direct questions, direct answers
How much does an owned system cost?
It depends on the process, integrations, migration and risks involved. A defensible range only comes after an initial understanding, so we do not publish a universal price.
How long does it take?
The timeline depends on scope and rollout risk. We organize the work in stages and plan a gradual transition when the operation cannot stop.
What about maintenance after delivery?
Software in production needs fixes, security updates, infrastructure and evolution. That cost belongs in the comparison, and continuity may stay with Sothen or move to another prepared team.
Is the code really mine?
Code ownership, licenses and access to data and infrastructure are defined in the contract before work begins. The goal is to avoid artificial dependency on Sothen.
What about migrating data from the current tools?
When migration is needed, it is sized in the scope: data availability, transformation, validation and transition strategy. Not every supplier provides the same export options.
Will my team accept switching tools?
It is the most underestimated part of the project, which is why design starts with the people who use it. A system built for your workflow usually gets easier adoption than a generic tool full of features nobody asked for.
Let us assess whether building makes sense
Tell us how the process works today, where it gets stuck and which tools are involved. The answer may be to build, integrate or keep what already exists.