One Supplier, Four Weeks of Float

Start with the float, not the probability. The quote is 16 weeks against a 20-week go-live, so the schedule float is 4 weeks. A four-week vendor slip does not merely delay the order - it consumes 100% of float, meaning any downstream installation, configuration or integration delay lands directly on go-live. Expected delay value is 0.3 x 4 = 1.2 weeks, which sounds tolerable until you note that the loss is not linear: at a 4-week slip you lose all absorbency, and the tail is what hurts.

Separate controllable from uncontrollable. The vendor's component allocation is outside your control. The date you place the order is entirely inside it - and every week of delay in ordering reduces the 4-week float one-for-one, converting a manageable risk into a near-certain slip. That asymmetry drives the first decision.

The contractual position is weak by design. A 2% credit capped at 6% transfers almost none of the commercial risk; on a large hardware order 6% may not cover the cost of a delayed go-live, and there is no cancellation-for-convenience right, so you cannot walk away if the vendor slips.

What the probability is worth. The 30% figure is verbal, unwritten and from an interested party. Treat it as a starting point for your own analysis rather than a validated input: ask for the component allocation confirmation, the production slot and the ship date in writing, and if those are refused, treat the probability as higher than 30%, because refusal to commit is itself information.

Test whether the mitigation can actually protect this date. The second source needs 8 weeks of qualification, then its own lead time, and it requires a segmentation redesign. Qualifying plus ordering cannot produce an appliance inside the 20-week go-live window. Therefore the second source cannot mitigate this project's schedule risk - presenting it to the steering committee as protection for this date would be misleading.

So evaluate it as what it is: a strategic option, not a mitigation. The question becomes whether the option value exceeds USD 120,000 plus the design-change cost. That value includes the next hardware phase, the negotiating leverage against the incumbent at the next refresh, and removal of a single-point dependency that has already shown weak contractual cover. Those are real benefits - but they belong to a portfolio or enterprise-risk decision, funded from a strategic budget line, not from this project's contingency.

What to do about this date instead:

• Place the order now. Delaying ordering burns the only controllable float.
• Convert the verbal warning into contractual content: written confirmation of component allocation, production slot and ship date, with milestone payments tied to those dates rather than to order placement.
• Seek a reschedule or cancellation right even at a price; the absence of a walk-away option is the structural weakness.
• Plan a partial-deployment fallback - install with the legacy segmentation for non-critical sites so go-live is not a single all-or-nothing event.

Recommended position to the committee: order now, negotiate written commitments and protections, and fund second-source qualification separately and explicitly as an enterprise hedge, with the honest statement that it will not save this date.

Allocate by type of uncertainty, not by convenience.

• Schedule reserve: the 4 weeks of float between the 16-week lead time and the 20-week go-live. This is already embedded in the plan and should be tracked as reserve, not treated as free time for other activities. Protect it explicitly - no one may consume it for design changes or scope.
• Contingency reserve: held for this identified risk (vendor slip) and other known risks in the register. The 30-day management contingency the committee is holding is best treated as management reserve if it is genuinely for unknown-unknowns such as a customs hold, port disruption or a component allocation being pulled after order.
• The distinction matters in reporting: if a known vendor slip is paid from management reserve, the project appears to have consumed reserve for an unidentified event, and the governance record misrepresents what happened.

Triggers, each with an owner and a date:

1. Written component allocation confirmation not received within 14 days of order - escalate to vendor executive sponsor.
2. Any published or verbal slip beyond 5 working days - activate expedite options (air freight, partial shipment from another region, loan stock).
3. Ship date confirmed later than week 14 - trigger the partial-deployment fallback and re-sequence installation.
4. Second slip notification - convene the steering committee with a revised go-live recommendation rather than absorbing it in weekly status reports.

The discipline point. A risk that is monitored weekly without a pre-agreed trigger simply becomes a surprise with a paper trail. Define the trigger, the action and the person who acts before the slip happens, while everyone is still reasonable about it.