When an infrastructure fund invests in a data center operator, it expects regular reporting that is clear and comparable from site to site. In practice this is rarely the case at the start: each site, especially an acquired one, calculates its indicators its own way. One site's PUE is not measured at the same point as its neighbour's, availability does not count the same events, and some meters are missing. The result: reassuring figures that do not support decisions.
DOES YOUR SITUATION LOOK LIKE ONE OF THESE?
1. WHAT INVESTORS WANT TO KNOW
Behind the tables, three questions: is service being delivered, are risks under control, and will sellable capacity keep up with demand?
Risk: near misses, maintenance completed versus planned, equipment at end of life
Capacity: installed, sold and actually used power, and available margin per hall
Energy and environment: consumption, PUE, water, refrigerants
Projects: progress of extensions and investments, cost and schedule variances
2. THE TRAP: FIGURES THAT DO NOT COMPARE
PUE depends on where it is measured, over what period and what is included. Two sites can show the same figure with very different performance. See what a point of PUE is worth.
Availability varies depending on whether planned maintenance, partial outages and incidents without customer impact are counted.
Capacity often mixes contracted power with power actually available, while it is the latter that limits sales.
The answer is not a tool but a rule: for each indicator, a written definition, a calculation method, a data source and an owner.
3. THE LINK WITH CONTRACTS AND REPORTING OBLIGATIONS
Internal indicators must be consistent with the service commitments signed with customers: a site that meets its internal targets but not its contractual commitments is exposed to penalties. They should also feed reporting obligations: energy use under France's tertiary energy-saving scheme and, for data centers of at least 500 kW, the annual declaration to the European database. One reliable data set is better than three reports built separately.
4. THE METHOD
2. Write the definition and calculation method for each
3. Check, site by site, the available data and missing meters
4. Align with customer contracts and regulatory reporting
5. Produce a first dashboard, then refine it over three or four months
This is one part of our post-acquisition technical integration offer, with our operations and energy experts. See also the typical situation "Our shareholders want a monthly dashboard common to all sites".
This article is general information. The reporting obligations mentioned should be checked in their current version for each site.
FREQUENTLY ASKED QUESTIONS
What indicators does a fund expect from a data center operator?
Service indicators (availability, incidents), risk (maintenance, end-of-life equipment), capacity (sold and available power), energy (consumption, PUE) and project progress.
Why are two sites' PUE figures not comparable?
Because they may be measured at different points, over different periods and with different boundaries. A common, written calculation method is needed.
How many indicators should be tracked?
A dozen well-defined ones are better than fifty poorly measured. What matters is that each has a definition, a method, a source and an owner.
Can investor reporting feed regulatory reporting?
Yes, and it should: one reliable data set can feed internal reporting, customer commitments and energy consumption and efficiency declarations.