What Data Do You Need for CAPEX Optimization? The StratePlan Setup Explained Simply

A common question before implementing Decision Intelligence is:

“Is our data even ready for mathematical CAPEX optimization?”

The answer is often simpler than expected.

To get started with StratePlan, a company needs neither a new ERP system nor a data project that takes years to complete.

In the simplest case, three pieces of information are sufficient to start:

Project ID → Expenses → Revenue

The decision-making model can then be expanded step by step based on this.

The key is not to collect as much data as possible.

The key is to link the data relevant to the investment decision.

Table of Contents

• 1. The Minimum Dataset
• 2. The Project ID as a Link
• 3. Merging Existing Data Silos
• 4. What Additional Data Can Be Considered?
• 5. Dependencies and Constraints
• 6. Defining Success Criteria
• 7. The StratePlan Setup in Five Steps
• Why the initial setup should be intentionally streamlined

1. The Minimum Dataset

For an initial optimization, the data structure doesn’t have to be complex.

In the simplest model, StratePlan initially requires the following for each investment project:

Project ID - Expenses - Revenues

For example:

Project 4711 - Investment / Expenditures: 10 million EUR - Expected Economic Contribution (Revenues): 15 million EUR

This already raises a fundamental mathematical question:

Which combination of existing projects generates the highest defined ROI / benefit within the available budget?

That is the foundation.

Additional data is added only when it is relevant to the specific decision.

2. The Project ID as a Link

In practice, the relevant information is often already available within the company.

The problem:

It is scattered across different systems and organizational silos.

For example, the CAPEX list contains the project and the investment amount/expenses.

Finance has profit contributions, cash flows, or NPV.

Operations knows the resource requirements and technical dependencies.

Business units have market, revenue, or capacity assumptions.

Management defines strategic objectives.

This information does not necessarily have to be migrated to a completely new system.

First, you need a common link.

This connection is the Project ID.

3. Consolidate existing data silos

The basic data model can therefore be represented very simply:

CAPEX / PPM + Finance / Controlling + Sales / Business Unit + Operations + Strategy / Management → Project ID → StratePlan

Existing ERP, Finance, PPM, Excel, or database structures can remain in place.

StratePlan operates at the decision-making level.

The key is that the information relevant to a project can be linked together via a unique ID.

This transforms distributed corporate data into a shared decision-making space.

The data does not have to be generated in the same location. It must be possible to unambiguously assign it to the same project.

4. What additional data can be included?

Based on the minimum dataset, the model can be expanded to reflect the actual business situation.

This may include, for example:

Costs
Profit
Cash flow
NPV
Revenue
Output
Duration
Resource Requirements
Capacities
Locations
Business Units
Project Categories

Not every company needs all of this information.

And not every investment decision requires the same data structure.

That’s why data preparation should be based on the decision-making question.

What information actually influences the decision?

Only this data needs to be included in the respective model.

5. Dependencies and Constraints

In real-world investment portfolios, projects can rarely be considered completely independently of one another.

Therefore, additional dependencies and constraints can be defined.

For example:

Project B can only be implemented if Project A is implemented.
Project C and Project D are mutually exclusive.
Project E must be implemented.
Only a limited budget is available for a specific location.
A certain resource is available only in limited quantities.
A business unit must be considered with a defined minimum volume.

These conditions mathematically represent the company’s reality.

StratePlan does not simply seek the theoretically highest return.

It seeks the best permissible combination of projects within the defined conditions.

6. Define Success Criteria (Public Sector)

Before optimization begins, it must be clear what “better” means for the specific decision.

The company therefore defines the relevant success criteria.

Depending on the use case, this could include, for example:

Maximizing the economic contribution
Maximizing NPV
Maximizing cash flow
Maximizing output
Optimizing strategic targets

The technology does not decide on its own what is important for the company.

Management defines the objective. StratePlan calculates the project combinations within the specified conditions.

This means that the strategic decision remains with management.

7. The StratePlan Setup in Five Steps

The basic setup can thus be broken down into five steps:

Step 1: Identify projects
Which investment projects should be included in the decision?

Step 2: Standardize Project IDs
Each project is assigned a unique identifier that can be used to link information from different data sources.

Step 3: Assign investments (expenditures) and revenues (business) or benefits (public sector)
For each project, at least one of the following is recorded: investment/expenditures or expected economic benefit, revenue/income, or—in the public sector—a defined benefit (0.01 to 0.99 or 1–15).

Step 4: Add Relevant Constraints and Dependencies
Budget, resources, dependencies, and other real-world conditions are defined.

Step 5: Define success criteria
Management defines which target metric is to be optimized.

This establishes the basic data structure for the calculation.

Why the initial approach should be deliberately lean

Decision Intelligence shouldn’t start with the question:

“How do we get all our company data into a new system?”

But rather with:

“What data do we need to better calculate this specific decision?”

That’s a fundamental difference.

A company doesn’t have to transform its entire data landscape first in order to begin using mathematical optimization.

It can start with a clearly defined decision and a lean dataset.

Project ID → Expenses → Revenue

After that, only the data, dependencies, and constraints needed for the actual decision are added.

Step by step, this creates a mathematically calculable decision space from existing company data.

And that’s exactly why getting started with decision intelligence can be much more streamlined than many companies initially expect.

Start with the decision. Not with the system.

Decision Intelligence for All.

Don’t take our word for it. Calculate it.