87 Days of Schedule Recovery Found Where the Critical Path Wasn't Looking

Industry Data Center

Capability Version Comparison & Change Tracking

The critical path had improved. The finish date had gotten worse. Between two schedule updates on a large data center project, 4,857 tasks had changed and something wasn't adding up. Nodes & Links compared every change, found that activities outside the critical path had quietly consumed all their buffer, and identified MV switchgear energization as the true controlling constraint. 87 days of recovery was unlocked immediately.
slip explained
0 -day finish
A set of activities outside the critical path quietly consumed all their buffer and started driving the finish date back
schedule recovery
Up to 0 days of
The curtain wall was the biggest culprit: fix it, and 87 days of schedule recovery open up immediately

Minutes,
not days

Comparing 4,857 changed tasks manually would have taken days. Nodes & Links did it in minutes

The Challenge

Between two consecutive schedule updates on a large data center project, the team had made real progress. The critical path had been compressed by 15 days. But the finish date had moved 47 days later. Something didn't add up. Every day a data center misses its go-live date costs the operator significant revenue, so understanding what was actually driving the slip, and fast, wasn't optional.

What was asked

The critical path is shorter but the finish date has slipped 47 days. What changed and where is the recovery opportunity?

How helped

Nodes & Links compared both schedule versions across all 4,857 modified tasks and found the real problem: it wasn’t on the critical path at all. A set of activities that had appeared safe, with buffer to spare, had quietly used up all that buffer and were now pushing the finish date back. The biggest culprit was the curtain wall, the building’s external cladding. 

Until the facade is weathertight, none of the internal fit-out or systems testing can begin, making it the true controlling constraint even though it never showed up as critical. Two further delay clusters were identified: the HV Switchroom (47 days) and the Diesel Generator systems (19 days).

The Result

The contradiction was explained and the 87-day curtain wall recovery opportunity was immediately actionable. The team had a complete picture of every delay driver across the program, delivered in minutes rather than the days a manual comparison would have taken.

Download Case Study

Enter your details below to receive your copy.

Nodes & Links hates spam, and we will never share your details with third parties.