Current guide
From pilot to operations: scaling municipal welfare technology in 2026
Many municipalities succeed in pilots but stop before full operation. Here is a practical model for scaling welfare technology and digital keys.

Why do pilots stall?
Pilots often have dedicated resources and limited scope. Scaling introduces more locations, roles and exceptions. Common obstacles are unclear ownership, manual steps, weak integrations, missing benefit measurement and technology that does not suit every user or door.
Start with the outcome
“Implement digital keys” describes a deliverable. An outcome describes the value, such as reducing key logistics by 15 minutes per shift or halving the time from a need arising to active access.
Standardise the workflow
Describe the normal process from registering users and locks to assigning, using, reviewing and ending access. The more local special processes there are, the harder rollout becomes.
Choose a platform that handles variation
Municipalities often have existing and new locks, mobile users and users who need fobs or cards, as well as locations with varying coverage. Administration should be shared even when the technology varies.
Prioritise integrations by impact
Entra ID, HR, electronic patient records, planning and response systems may be relevant. Start with integrations that remove the most manual work or risk.
Make benefit realisation part of operations
The Norwegian Directorate of Health emphasises management engagement. Link each benefit to a baseline, target, data source, owner and measurement frequency.
Five steps from pilot to scale
- Define scope: Choose one area with clear volume.
- Stabilise: Remove errors and document procedures.
- Repeat: Roll out to a new area using the same method.
- Integrate: Automate the most important manual steps.
- Manage: Transfer responsibility, budgets and measurement into the operational organisation.
Conclusion
Scaling succeeds when technology is part of a complete service model. Municipalities need clear outcomes, ownership and a repeatable method more than an ever-larger pilot.
Frequently asked questions
How long should the pilot last?
Long enough to cover normal shifts, user changes, errors and variation in coverage.
Must all locks be replaced at once?
Usually not. Phased rollout reduces risk.
When should integrations be developed?
When the workflow is understood and it is clear which manual work should be removed.