Buying a data center means buying its history: equipment installed at different times, maintenance contractors chosen by someone else, procedures that may or may not be written down, a team with its own habits. After two or three acquisitions, an operator ends up with as many ways of running sites as it has sites. The temptation is to impose a single model very quickly. That is often when incidents happen.
DOES YOUR SITUATION LOOK LIKE ONE OF THESE?
1. THE PROBLEM: INVISIBLE RISKS
As long as each site is run its own way, nobody really knows which one carries the most risk. A UPS at end of life on one site, a maintenance contract that does not cover the chillers on another, generator tests no longer carried out on a third: any of these can cause an outage. And the figures reported to shareholders, calculated differently, do not reveal them.
2. MISTAKES TO AVOID
Changing everything at once. Replacing contractors, monitoring and procedures simultaneously on a live site multiplies the chances of human error, one of the main causes of data center incidents.
Standardising on paper. A group procedure that ignores a site's technical specifics will not be applied, or will be applied badly.
Forgetting the teams. Technicians at an acquired site know its weaknesses better than anyone. Sidelining them means losing that knowledge.
Investing before knowing. Starting works on the most visible site rather than the riskiest one.
3. A FOUR-STEP METHOD
2. A risk ranking, from most critical to least urgent, and an action plan at 3, 6 and 12 months
3. Gradual harmonisation: maintenance contracts at renewal, procedures built on existing best practice, common indicators
4. Follow-up: a common dashboard and regular reviews with management and shareholders
Order matters. The review comes before decisions, and changes are made site by site, outside high-risk periods such as extension works or customer audits.
4. THE ISSUES THAT ALWAYS COME UP
Documentation. Incomplete as-built files, outdated single-line diagrams: this is the first workstream, and the least visible.
Maintenance. Contracts with different scopes and response times. Their renewal date is the right moment to harmonise.
Density. Customers want more power per rack. Each site has a different electrical and cooling margin: you need to know it before selling it.
Regulatory obligations. Energy reporting, efficiency, refrigerants: each acquired site arrives with its own situation, sometimes behind schedule.
5. WHO DOES IT
An operations team busy with day-to-day running rarely has time for this work. An outside view, independent of incumbent contractors, makes it possible to assess the sites objectively and avoid debates about habits. This is the purpose of our post-acquisition technical integration offer, with our design, commissioning and operations expert. See also the typical situation "We acquired three sites with different maintenance practices".
This article is general information. The reporting obligations mentioned should be checked in their current version for each site.
FREQUENTLY ASKED QUESTIONS
Where should you start after acquiring several data centers?
With a technical review of each site using a common grid, then a risk ranking. Investment and harmonisation decisions come next.
Should maintenance contractors be changed right away?
Rarely. It is better to harmonise contracts at renewal, site by site, to limit the risk of errors during the transition.
What is the main incident risk during an integration?
Simultaneous changes on a live site, which multiply the risk of human error, one of the main causes of data center incidents.
Is an external review useful if the team knows its sites?
Yes, to compare sites against the same grid and set priorities objectively, independently of incumbent contractors and local habits.