5 min read. For project managers who keep discovering handover problems in the weekly meeting.
Key takeaways
- Delay rarely lives inside a trade’s scope. It lives in the handovers between trades that nobody planned.
- An interface is only managed when it exists in the programme as a milestone with a named owner and a date.
- A one-day interface workshop early in the project is the cheapest delay mitigation a data center can buy.
Ask any data center delivery team where their last three weeks of delay came from and the answers are rarely inside a single contract. The busbar was ready but the ceiling grid was in the way. The chilled water system was flushed but the controls points were not terminated. The generators arrived on programme and waited five weeks for their cabling. Each trade did its job. The project still slipped, because the work between the trades belonged to nobody.
Why data centers are interface-dense
A data center packs an unusual number of systems into a small footprint on a fast programme: power distribution, cooling, fire suppression, controls, security and networks, all converging on the same rooms in the same weeks. Add the commissioning chain, where every level depends on completions from several trades at once, and the interface count grows geometrically. On a typical hyperscale building the genuinely risky handovers number in the dozens, and almost none of them appear in anyone’s contract programme.
From minutes to milestones
Most projects manage interfaces in meeting minutes: someone raises the busbar and ceiling question, an action is noted, and the minute dies in a folder. The fix is structural, not heroic. Every significant handover becomes a milestone in the master programme, logic-linked to the activities on both sides, with a named owner from each trade. Now the interface behaves like every other piece of scope: it has float, it can go critical, it shows up in the lookahead, and slippage is visible weeks ahead instead of being discovered in the corridor.
Each trade did its job. The project still slipped, because the work between the trades belonged to nobody.
The one-day workshop that pays for itself
The practical route in: put the MEP contractors, the commissioning manager and the planner in one room early, walk the sequence area by area, and list every point where one party needs something from another to proceed. Score each interface for programme risk, then write the risky ones into the baseline as owned, dated milestones. On a multi-hundred-million build, one structured day and a few dozen milestones routinely prevent months of accumulated drift. It is the cheapest delay mitigation available on a mission-critical project.
How many unowned interfaces are in your programme?
Our fixed-fee Schedule Health Check maps them, and our programme consultancy puts them under control. Get in touch.
