How a mid-market plant engineering company with 120 employees and 15 parallel projects moved from five fragmented truths to a learning Operating System – and today proactively manages projects instead of merely coordinating them.
Current Situation
The company plans and builds technical systems for industrial clients – projects lasting from six to eighteen months, involving multiple trades, subcontractors, and a high density of coordination. This is a business where dozens of transitions occur between bidding and final billing.
The ERP system knew the order positions. The construction manager's Excel knew the reality on site. The filing system knew the emails regarding changes. And the project team filled in the gaps in between – anew every day.
The operational reality:
- What was in the performance specification, what happened on site, and what ended up in the final invoice was tracked in separate lists – merging them took hours.
- Deadlines, capacities, and sign-offs with subcontractors depended on individual conversations and loose documentation – without a common picture, gaps, duplicate work, and disputes over status and responsibility arose.
- Personnel, equipment, materials, and external services had to fit across multiple sites simultaneously – the ERP saw order positions but not the daily operations.
- Progress, quantities, delays, and special cases ended up in construction manager's spreadsheets because the ERP was too rigid – the longer this persisted, the more truth drifted away from the core system.
- Approvals, deviations, and documentation were managed via email and Drive – what happened operationally often arrived late or incomplete in the project file.
The problem was not in a single module.
No system provided the operational overview – and the project team pieced together complexity anew every day. This worked as long as the project managers were experienced enough and too many projects weren't running simultaneously. However, more projects meant exponentially more coordination.
The margins did not increase – but the friction did.
Three Stages to Operational Freedom
We structured the project along our 3-stage model. Each stage builds on the previous one. After 10 months, all three stages are running productively.
| Stage 1 – Order | Stage 2 – Automation | Stage 3 – Freedom |
|---|---|---|
| Consolidating projects, resources, subcontractors, orders, and documentation into one system | End-to-end processes from quotation to final invoice, changes integrated | AI-based risk detection, predictive planning, automated estimation, early warning systems |
Stage 1 – Order
Before we could automate processes, we needed to make visible where the operational truth actually lay – and where it was lost. In Stage 1, we established:
- A mapping of all project data – we clarified what comes from the ERP, what lives in the construction manager's Excel, what is contained in email archives, and where subcontractor agreements are documented.
- A central project model – one project, one status, with order positions, resources, subcontractor status, deadlines, and risks in one place.
- Clear transfer logic – when is a service considered ordered, when started, when accepted, who documents what and where.
- The integration of the existing ERP for estimation and order data – no parallel system, but an operational layer above it.
The most challenging phase was not the technology. It was determining what constitutes a binding project status. Until then, each participant had their own: the project manager had their Excel, the construction manager their construction log, accounting had their ERP positions, and the subcontractor had their last email. We collaboratively translated these different realities into a model – not just technically but also organizationally.
The hardest realization:
Three project managers, asked about the same project, cited three different completion percentages. Not because they were wrong – but because each was looking at a different data source.
The implementation proceeded step by step: workshops with project management, construction managers, procurement, and accounting. We intentionally piloted with an ongoing mid-sized project – not a new one, but one that already had the typical problems. Key users from project management took on the role of multipliers. After four months, the first three projects were running in the new system. After seven months, all active ones.
Stage 2 – Automation
Based on the new structural clarity, we specifically added automation. The company saw the greatest impact in change order processing: Previously, the construction manager would identify a deviation, note it in their Excel, and send an email to the project manager. The project manager checked it and forwarded it to estimation. Estimation needed quantities from another spreadsheet. Accounting required approval from a third email. By the time the change order reached the client, weeks had passed – and the documentation was incomplete.
Today, the construction manager records the deviation directly in the system – with quantities, photos, and references to the order position. The project manager sees the change order immediately, reviews, and approves it. Estimation already has all data structured. Accounting prepares the final invoice based on comprehensive performance records, not based on gathered emails. What used to take weeks of follow-up has become an integrated process that runs parallel to execution.
Further automations in this phase:
- Subcontractor transfers with uniform status and structured documentation – fewer queries, clearer responsibilities.
- Resource allocation across projects – who is scheduled where, what equipment is on which site, where bottlenecks arise.
- Status changes on site automatically update the project status – project managers and management are looking at the same status.
- Digital recording of inspections and approvals instead of email communication and scattered folder structures.
After Stage 2, operational business ran more smoothly, accurately, and quickly. However, strategic questions – which projects pose risks, where resources are really tight, how margins evolve across the portfolio – remained with project managers and management.
Stage 3 – Freedom
With a stable data foundation and functional automation, the system reached a point where it can co-responsibly handle strategic questions. Stage 3 has transformed the way the company manages projects.
AI-Powered Risk Detection
The system identifies project patterns that have previously led to delays or margin losses – and reports them weeks before they escalate. A specific trade that delays the start of work. A subcontractor who repeatedly submitted change orders on similar projects. A resource configuration historically leading to bottlenecks. Project managers see risks early – and have time to take action before they escalate into crises.
Forecast-Based Resource Planning
The system plans personnel, equipment, and external services across all ongoing and anticipated projects. It predicts where bottlenecks will arise in six to eight weeks, based on project progress, seasonal patterns, and typical deviations. Management proactively reallocates resources – instead of renegotiating each site individually when it’s too late.
Automated Quotation Estimation
New quotations are based on real data from comparable projects. The system knows the actual expenditures, change order rates, typical delays, and margin realities of similar contracts. The estimator adjusts the system-generated quotation – rather than building a new estimation from scratch. Quotations are created faster, align closely with reality, and safeguard the margin.
Early Warning Systems for Change Orders and Subcontractors
The system identifies change order risks before the construction manager reports them – for example, when quantities exceed the performance specification or when deviations from estimation show a critical pattern. The same applies to subcontractors: Reliability, quality, and delivery adherence are continuously evaluated. Those who generate risks in projects will show up in the next awarding decisions accordingly.
Measurable Results After All Three Stages
| Area | Before | After Stage 2 | After Stage 3 |
|---|---|---|---|
| Project Execution | Progress tracked via tools, spreadsheets, and queries | An operational picture per project in one place | Project portfolio manageable as a whole, risks visible early |
| Resource Management | Capacities adjusted manually | Transparency over teams, equipment, external services | Forecast-based across all projects, 6–8 weeks in advance |
| Subcontractors | Individual conversations and loose documents | Uniform transfers and status | Continuous evaluation factored into awarding decisions |
| Change Orders & Billing | Late and incomplete data | Capture during the process, comprehensive evidence | Early warning for change order risks before they arise |
| Quotation Estimation | Manual, based on experience | Structured, based on current master data | Data-driven on real project trajectories of comparable contracts |
| Control Mode | Reactive – chasing after | Proactive – with early signals | Predictive – the system contributes to decision-making |
What Has Changed for All Parties Involved
The system has transformed the way the company conducts projects. Not only operationally – but also strategically. Project managers see the operational status of their projects in one place. They spend time on management instead of compiling information and recognize risks before they escalate. Construction managers record progress, deviations, and change orders directly in the system – what happens on site is immediately reflected in the project file. Subcontractors work with uniform transfers and clear statuses; disputes over status and responsibility have significantly decreased because everyone is looking at the same truth. Management sees, across all projects, where margins are generated and where they erode – and reallocates resources before bottlenecks jeopardize projects. The sales team estimates new projects based on real data from comparable contracts, not gut feelings.
Earlier, the day was filled with coordination. Now it’s filled with decision-making.
And Perhaps Most Importantly
More projects no longer make the organization more confusing. The company has decoupled its growth from its coordination efforts – the number of projects can increase without the number of coordination meetings growing alongside it.
What This Project Demonstrates
This project showcases a pattern common in project-based companies: The ERP knows the order positions, the Excel knows the on-site reality, the filing system knows the emails regarding change orders. And the project team fills the gaps in between – every day, in every project. The first step was not a better ERP or a new planning tool. It was the decision to define a binding project status that all stakeholders could trust – and to make this status operationally accessible, not just in reporting. Only this order enabled automation. Only this automation enabled freedom.
Order enables automation. Automation enables freedom.
For those managing multiple projects simultaneously who notice that growth generates more coordination rather than more margin, a crucial question must be asked: Do all involved parties see the same picture of ongoing projects – or is everyone constructing their own? If the answer is "everyone is building their own," the leverage does not lie in an additional tool. It lies in a shared operational model – and in what the company can build upon that once it is established.