Before purchasing or building another system, examine the work itself. A process may need to be simplified, delegated, automated, or eliminated rather than preserved inside new technology. The right decision depends on value, frequency, predictability, risk, required judgment, scale, and strategic importance. Technology should support a better process, not make an outdated one more permanent.
Organizations often begin technology discussions with a solution already in mind. We need a new system. We need an application. We need AI. We need to digitize this form.
A better opening question is: What should happen to this work?
In one organization, an enterprise resource planning system had been heavily customized around existing workflows. Some of those workflows were burdensome and unnecessarily complex, but the software preserved them in code. Every customization created more technical debt, made updates harder, and tied the organization more tightly to processes it should have reconsidered.
Technology can make work faster. It can also make bad work harder to change.
Stop before you improve
The first option should always be elimination.
Organizations accumulate reports, approvals, meetings, forms, and recurring tasks because they once served a purpose. Over time, the purpose changes or disappears while the activity remains. Stopping the work feels uncomfortable because completion has become evidence of responsibility.
Ask:
- Who uses the output?
- What decision does it support?
- What would happen if it were not completed?
- Could it occur less often?
- Is the effort proportional to the value?
If no one can explain why the task matters, do not automate it. Stop doing it and see what happens.
Simplify before you digitize
We were once asked to digitize a human-resources form involving complex calculations and numerous special situations. Reproducing every rule would have required substantial custom development.
The deeper question was whether the form needed that degree of precision. By creating reasonable buckets of choices instead of calculating everything to the penny, the process became easier for employees and administrators. It could then be built with an ordinary form tool rather than bespoke software.
The technology problem was solved by redesigning the process.
Simplification may mean reducing choices, removing unnecessary approvals, standardizing exceptions, clarifying ownership, or accepting an appropriate level of precision. A process does not become modern merely because it moves from paper to a screen.
Delegate when the work can develop someone
Delegation is difficult because experienced employees often believe they can complete the task faster or better themselves. That may be true today. It may not remain true if no one else is ever given the opportunity to learn.
Delegation does not mean abandoning responsibility. It means giving someone else the next level of execution, along with context, authority, support, and accountability.
A task may be a burden to a senior leader and a valuable development opportunity for a newer employee. Holding onto it can limit both people. The leader remains overwhelmed, while the employee is denied the chance to build capability.
Good delegation asks:
- Is this work appropriate for another role?
- What does the person need to know before beginning?
- Which decisions can they make independently?
- When should they check in?
- What does a successful result look like?
The goal is not simply to move work. It is to build capacity.
Automate predictable work
If a process is repetitive, documented, rules-based, and does not require much judgment, automation should usually be considered before AI.
A workflow can move information, send notifications, create records, apply consistent rules, route approvals, and identify exceptions without introducing unnecessary variability. Free and low-cost automation tools can handle many processes that once required manual effort or custom development.
AI may still contribute. It could classify an unusual request, summarize a document, or provide analysis inside an otherwise predictable workflow. The automation manages the process. AI assists only where interpretation adds value.
This approach limits cost and risk while preserving consistency.
Buy when the need is common
Buying an established product often makes sense when many organizations share the same need. Payroll, accounting, customer relationship management, document storage, and project management are rarely unique enough to justify building from scratch.
The difficult part is resisting the urge to reproduce every existing practice inside the new system. A product may represent a more mature process than the one currently used. Before requesting customization, ask whether the organization should adapt its workflow instead.
Requirements still matter. Evaluate security, integration, support, total cost, accessibility, usability, and the ability to retrieve your data. Also distinguish genuine requirements from preferences based on familiarity.
“We have always done it this way” is not a system requirement.
Build when the capability creates distinction
Custom development is appropriate when the capability is strategically important, commercially differentiating, or impossible to achieve responsibly with available products.
Building offers control but also creates obligations. Someone must maintain the code, manage security, support users, update integrations, document the system, and continue funding it after the initial excitement fades.
The question is not only whether the organization can build something. It is whether it should own that capability over time.
Before building, clarify:
- What unique value will this create?
- Why can an existing product not meet the need?
- Who will maintain and support it?
- How will success be measured?
- What is the exit strategy if priorities change?
Make the decision in the right order
The title may begin with building, but the decision process should usually move in the opposite direction:
- Stop work that no longer creates value.
- Simplify the work that remains.
- Delegate work that can develop capacity.
- Automate predictable steps.
- Buy mature capabilities that already exist.
- Build only where ownership creates meaningful advantage.
This sequence prevents organizations from spending money to preserve complexity. It also makes technology decisions part of organizational design rather than isolated purchasing decisions.
The best technology project may be a new platform. It may also be a smaller form, a clearer role, an automated workflow, or permission to stop doing something that no longer matters.
Put the idea to work
### What could this make possible for your organization?
Bring Foster + Foundry the challenge, opportunity, or question. We will help clarify a practical next step.
