Data Centres Are in Demand. But Before Investing, Answer These Five Questions.

AI, cloud adoption and the expansion of digital infrastructure are creating significant demand for computing capacity.

But a strong investment theme does not automatically make every data centre project investable.

Before focusing on the projected IRR, investors should be able to answer five fundamental questions.

1. Is the power truly secured?

Power capacity is often presented as a headline number: 20 MW, 50 MW or more.

But investors need to distinguish between power that has been requested, allocated, contracted and is actually deliverable within the investment period.

The relevant questions are:

• How much capacity is contractually secured?

• When will it become available?

• At what cost?

• Can the site accommodate future expansion?

In several European markets, the scarce asset is not the building or the land. It is access to deliverable power.

2. Who will pay for the capacity?

A signed contract with a creditworthy customer is fundamentally different from a pipeline of “advanced discussions”.

Investors should examine:

• the quality and concentration of the customers;

• the duration and structure of the contracts;

• the pricing and indexation mechanisms;

• the guarantees supporting the obligations.

Demand for data centres may be structural. Revenue at the individual asset level is not automatically secured.

3. What risk is the investor actually taking?

“Data centre investment” can describe very different exposures:

• an operational, income-producing facility;

• a fully permitted but unbuilt project;

• a development awaiting power or customers;

• an early-stage site supported primarily by its future potential.

Each belongs to a different risk category.

Investors must identify which risks remain—permitting, power, construction, financing, commissioning or commercialisation—and determine whether the expected return genuinely compensates for them.

4. How much capital is required before the asset produces stable income?

The initial equity requirement rarely tells the entire story.

Construction overruns, financing costs, equipment procurement, tenant-specific requirements and future expansion can materially increase the total capital needed.

The right question is not simply:

How much will it cost to build?

It is:

How much capital will be required before the asset reaches stabilised cash flow?

5. What will make the asset institutional at exit?

An exit strategy cannot rely solely on continued enthusiasm for AI and digital infrastructure.

A future institutional buyer will assess:

• contracted and diversified income;

• remaining contract duration;

• power availability and scalability;

• operating performance;

• future capital expenditure;

• the ability of the facility to remain commercially relevant.

The question is not only whether the data centre can be developed.

It is whether it can become an asset that long-term institutional capital will want to own.

Data centres remain one of the most compelling structural investment themes in private markets.

But sector growth does not eliminate asset-level risk.

The strongest opportunities are those where power, customers, execution and exit liquidity are aligned—not simply those carrying the AI label.