Challenging IBM findings line by line.

A draft is a position, not a bill.

The draft finding compares what IBM believes is deployed against what it believes you are entitled to. Both halves are built from the outside, from a data collection tool and from IBM's own records, and both tend to read in the vendor's favor. Treating the draft as a number to negotiate down concedes the framing. The buyer side move is to treat each line as a claim that has to be proven, then test it against the facts you control.

Wrong PVU values.

Processor Value Unit findings are the most common place to win. PVU is core-based: the PVU rating per core, for example roughly 70 PVU on a typical Intel core, multiplied by the cores allocated to the IBM software. Drafts routinely count against peak or total host capacity rather than the cores the software actually used, apply the wrong rating per core for the processor, or ignore that allocation changed over the period. Recalculating PVU from your own evidence frequently moves the number sharply.

Denied sub-capacity.

The largest single swing is usually sub-capacity. If IBM defaults you to full-capacity charging, the gap is enormous: WebSphere using 4 of 32 host cores is 480 PVU under sub-capacity against 3,840 PVU at full capacity. IBM denies sub-capacity when it believes a condition failed, the tool was missing, agents were broken, or reports were not retained. Each denial is challengeable on the facts, period by period. Where your ILMT data shows the conditions were met, the sub-capacity position is reinstated for that period.

• Confirm the eligible tool was deployed within 90 days of first deployment
• Show continuous operation and quarterly reports retained two years
• Apply the correct virtualization rule for each hypervisor in scope
• Dispute full-capacity defaults for periods the evidence actually covers

Missing entitlement offsets.

Drafts count deployments but quietly omit credits. Entitlements you already own, unused licenses elsewhere in the estate, bundled rights, and prior purchases that offset the claimed shortfall are often left out. Restoring those offsets reduces the net position before any commercial discussion begins. This is reconciliation, not negotiation: you are correcting the arithmetic, not asking for a favor.

What this means under audit.

Challenging line by line is how the Challenge step of Contain, Reconcile, Challenge, Settle actually works. The point is to close the gap on the facts before anyone talks money, because a finding corrected on evidence stays corrected, while a finding merely discounted can resurface. Once the wrong PVU, the denied sub-capacity, and the missing offsets are dealt with item by item, the settlement conversation starts from a defensible number rather than IBM's opening claim.