How a facility management company with 200+ employees moved from five tools to a centralized operating system – reducing daily scheduling time from three hours to 35 minutes.
Initial Situation
A facility management company oversees commercial properties, industrial sites, and public institutions – offering services from building cleaning to technical maintenance and security. Over 200 employees work varying shifts at different locations. Qualifications, property requirements, and special cases change daily.
Formally, the company was digitalized: Orders in the ERP, availability in Plantool, special cases in Excel, qualifications in the minds of individual schedulers. And the phone served as a universal fallback tool for anything that slipped through the cracks.
The operational reality:
- Scheduling took over three hours every morning
- Sick leave and special cases necessitated constant rescheduling
- Qualifications and property approvals were not centralized
- Communication occurred through phone chains – information arrived late and/or incomplete
- Growth did not create more freedom but rather increased coordination efforts The problem was not the individual tool. The problem was that five tools did not constitute a system. Scheduling worked as long as the most experienced people were available and nothing unexpected happened. But as the company grew, unexpected issues occurred more frequently.
The operational complexity had grown larger than the systems that were supposed to manage it.
Our Approach: First Order, Then Automation
The project followed our 3-phase model. Each phase builds on the previous one – without a foundation, no automation; without automation, no freedom.
| Phase 1 – Order | Phase 2 – Automation | Phase 3 – Freedom |
|---|---|---|
| Centralize deployment, personnel, and property data; connect existing tools | Automate scheduling, task logic, status updates | Forecasts, early warning systems, data-driven calculations |
Phase 1 – Order
Before we could think about automation, we had to make visible what was to be controlled. We analyzed the existing system landscape: What data exists where? What is reliable and what is outdated? Where is knowledge located in people instead of systems? What we accomplished:
- Consolidation of all order, property, and personnel data in a central application
- Establishing a qualification and approval logic – which employee is allowed to perform what activity at which property
- Defining clear status logic for deployments, shifts, and availabilities
- Integrating the existing ERP for order data – no parallel operation, but genuine integration
We consciously decided not to simply digitize data that was not maintained in any system – like informal agreements on deployment preferences or handwritten notes about property specifics. Instead, we examined which knowledge was truly operationally relevant, and only the relevant information was transitioned into the new structure.
The hardest part was not the technology. It was writing down for the first time what had previously existed only in the minds of the most experienced schedulers.
The implementation was gradual: workshops with dispatchers, property management, and team leaders, involvement of key users as multipliers, module-wise activation, temporary parallel operation for the most critical processes. After three months, planning was running in the new system. Not as a prototype, but in a productive environment.
Phase 2 – Automation
Based on structural clarity, we added targeted automations. What this means in practice is illustrated by an example from morning scheduling.
Before
Every morning began with a check-in: Who is sick? What special cases exist? Which replacement has the right qualifications and property approval? The scheduler would open the ERP, Plantool, two Excel spreadsheets, and start making phone calls. Three hours later, the daily plan was established – until the first sick leave at 9 AM.
After
The system already shows gaps when opened. Sick leaves are reported digitally and automatically trigger a proposal: Which available employees have the appropriate qualification, property approval, and shift compatibility? The scheduler reviews, confirms, or adjusts. Affected team leaders and property managers see the change immediately in the system.
From three hours of reactive planning, it has become 35 minutes of proactive management.
Examples of further automations that emerged in Phase 2 include:
- Status updates from teams trigger follow-up processes – handover, documentation, billing
- Qualification processes and property approvals are automatically monitored and escalated
- Rule-based task distribution for standard deployments
- Digital confirmation of deployment by employees on-site
Measurable Results
| Area | Before | After |
|---|---|---|
| Scheduling Time (daily) | approx. 3 hours | approx. 35 minutes |
| Follow-up Questions in Team | Dominated the day | reduced by approx. 80% |
| Qualification Checks | Manual, error-prone | System-supported, automatic |
| Data Status | Distributed across 5 systems | A single centralized software |
| Management Capacity | Reactive – after the fact | Proactive – with early warnings |
What Has Changed for Everyone Involved
The system has not only eased scheduling. It has changed the way property management, team leaders, and clients collaborate. Scheduling now plans ahead instead of behind, sees bottlenecks before they escalate, and spends time making decisions rather than cross-checking data. Team leaders on-site see their deployment plan digitally, report status directly, and receive changes in real-time rather than through phone chains. Property management and clients look at the same status – follow-up questions decrease, as all parties access the same operational truth.
The organization became freer. Not because there's less work – but because it runs in a structured manner.
Phase 3 – Freedom
With a stable data foundation and functioning automation, the point was reached where the system could begin not just to react – but to anticipate. Phase 3 fundamentally changed the nature of operational work.
Demand Forecasts Instead of Gut Feelings
Based on historical deployment data, seasonal patterns, vacation phases, and property-specific metrics, the system now predicts what qualifications will be needed when and where – up to eight weeks in advance. Staffing shortages become visible before they arise. Hiring new staff is no longer reactive after the first crisis but data-driven and planned.
Early Warning Systems
The system proactively reports when critical patterns become apparent: properties where sick leaves pile up. Qualifications that become scarce due to departures. Teams whose composition historically leads to errors. What was previously only visible during the next escalation phone call now appears as an indication in the system weeks in advance.
AI Agents for Standard Cases
The system now autonomously makes simple scheduling decisions. A sick leave at 5:30 AM automatically triggers the search for appropriate replacement. The first qualified, available, and approved employee is contacted – via app, with a confirmation request. Once confirmation is received, the deployment is rescheduled, property managers and team leaders are informed, and the shift continues. The scheduler now only checks in the morning what went wrong – instead of setting everything up themselves.
Data-Based Quotation Calculations
New properties are priced based on comparable deployments. The system understands the real effort of similar properties, the seasonal fluctuations, and the typical escalations. Quotes are generated in minutes instead of days – and reflect reality because they are based on it.
Measurable Results After All Three Phases
| Area | Before | After Phase 2 | After Phase 3 |
|---|---|---|---|
| Scheduling Time (daily) | approx. 3 hours | approx. 35 minutes | approx. 20 minutes, primarily for control |
| Planning Horizon | Hourly up-to-date, reactive | Daily up-to-date | Several weeks in advance |
| Follow-up Questions in Team | Dominated the day | approx. 60% reduced | approx. 85% reduced |
| Bottleneck Detection | When the absence occurs | The day of the absence | Weeks in advance |
| Quotation Calculations | Days per new property | Reduced to a few hours | Within hours, data-based |
| Management Mode | Reactive – after the fact | Proactive – with early warnings | Forward-looking – the system anticipates |
What Has Changed for Everyone Involved
The system has changed the way the company operates. Not just operationally – but also strategically. Scheduling is no longer a reactive center but a management role. Schedulers review, confirm, adjust – but the system carries the operational load. Team leaders on-site work with complete plans that include realistic buffers for standard deviations. Property management and clients now receive proactive information in many cases before they would have called: "We have identified a bottleneck in the maintenance appointments for next week; here is our suggestion." Management plans strategically – new locations, new services, new clients – based on reliable data, not on gut feelings.
In the past, operational reality dictated the day. Today, the company determines what operational reality should look like.
And Perhaps Most Importantly
The company has decoupled its size from its complexity. Growth no longer means more schedulers, more phone calls, or more chaos. Growth means more data points – and they make the system better with every deployment.
What This Project Demonstrates
This project demonstrates a pattern that is particularly common among service providers: operational complexity lies not in the individual order, but in the sum of shifts, qualifications, special cases, and constant rescheduling. No single standard tool captures that – resulting in workarounds, phone chains, and Excel spreadsheets. The first step was not more technology. It was the decision to transfer the knowledge residing in individual minds into a shared system. This may sound trivial, but it was the most challenging phase of the project. Only this order enabled automation. Only this automation enabled freedom.
Order enables automation. Automation enables freedom.
Anyone coordinating resources at varying locations daily should ask themselves: Do my tools create a system – or do they just form an assortment of individual solutions? If the answer to the second question is yes, the leverage is not in a better tool. It lies in a shared operational truth – and in what can be built upon it once it is established.