Warehouse System Launch at Northline Distributors
Northline Distributors is deploying a $2.4 million warehouse execution system across three distribution centers. The project is six weeks behind a 16-week plan and 10 weeks from peak season. The vendor's integration testing needs four weeks, followed by two weeks of defect fixes and one week of cutover. The project manager leads a matrix team: the IT integration developer is assigned 50% but was pulled onto a ransomware recovery effort; the operations lead refuses to release 6 hours per week of supervisor time for acceptance testing; and the controls engineer who knows the conveyor interfaces resigned with three weeks' notice. Eight of 22 acceptance tests have failed, mostly interface-related. Contingency is $120,000. The steering committee meets in 48 hours, and the CIO and COO are blaming each other. Go-live before peak is valued at $300,000 in avoided temporary labor.
1. Assess whether the project can go live before peak. Show schedule math and present at least three options with tradeoffs.
Read the worked answer
Remaining critical path: 4 weeks integration testing + 2 weeks defect fixes + 1 week cutover = 7 weeks. Peak is in 10 weeks, so the plan has 3 weeks of buffer only if the developer is effectively full-time and the eight failed tests close on schedule. At 50% allocation, testing alone stretches to about 8 weeks, consuming the buffer and pushing cutover past peak. Options: full go-live requires the developer at 100%, daily defect triage, and an acceptance-test freeze; value is $300,000 but risk is high. Phased go-live puts DC1 live in about 9 weeks and DC2/3 after peak; it may capture $100,000 to $150,000 of value and limits blast radius. Defer all and pay up to $300,000 for temporary labor; lowest technical risk but highest certain cost. Recommend phased unless the developer is formally reassigned and the eight failures drop to two or fewer by week 3. Three weeks of buffer cannot absorb both a ransomware-driven resource loss and a key-person resignation.
2. Recommend the actions and decisions for the 48-hour steering committee meeting, including use of the $120,000 contingency.
Read the worked answer
At the steering committee: require the CIO and COO to co-own one decision, not blame each other. Request the developer at 100% for three weeks or fund a contractor from contingency at $50,000 to $60,000. Secure operations supervisors for 6 hours per week each, with backfill. Retain the controls engineer as a part-time advisor for eight weeks at $15,000 to $20,000 and document the conveyor interfaces. Authorize up to $40,000 for test automation and overtime. Set a go/no-go date three weeks before peak. Use contingency for schedule-critical resources, not scope expansion. Publish a single RAID log and name one accountable project manager. If the developer cannot be released, choose phased go-live now rather than pretending the full plan is credible.
3. What no-go criteria should the steering committee set?
Read the worked answer
No-go if any safety or inventory-integrity test fails; if fewer than 18 of 22 acceptance tests pass by the go/no-go date; if the developer is below 100% allocation for two consecutive weeks; if rollback has not been rehearsed; or if DC1 cannot process a peak-day order volume in a load test. Also no-go if contingency spending would exceed $120,000 without new approval. A no-go should trigger the pre-approved phased or defer option, not a crisis. This converts the decision from a blame contest into a set of observable conditions the team can manage.