Contain. Reconcile. Challenge. Settle.

An IBM audit is won or lost in the first weeks, before any number is even calculated. The moment you return a full estate export, you have handed IBM the basis for its claim and lost control of the clock. Containment means scoping the data request down to what the audit clause actually entitles IBM to see, curating what is sent, and holding everything else until your own position is built.

This is also where timing becomes leverage. The audit calendar, the support renewal date and IBM's own quarter end all move the negotiation. Controlled early, they work for you. Left open, they work against you.

Before IBM presents a finding, we calculate the answer independently. That means recounting PVU against the correct per core table, rebuilding the sub-capacity evidence from ILMT, and netting entitlements that IBM's opening position routinely leaves out. The point is to reach the audit conversation already holding a defensible number, so the discussion starts from your evidence rather than IBM's assumptions.

Most of the reduction comes here. IBM findings are built on a stack of assumptions, and each one is a place to push: incorrect PVU per core values, sub-capacity denied over an ILMT gap that was narrower than claimed, full capacity applied where sub-capacity rights hold, entitlement offsets ignored in the netting. Challenged line by line with evidence, those assumptions give way. On average, a disciplined challenge lands 30 to 50% of findings, and a strong position carries further.

The last step is where a one time reduction becomes a durable outcome. We negotiate the corrected number down and use the settlement to fix what comes next: forward renewal terms, sub-capacity reinstatement, and language that closes the gap so the same finding cannot reopen at the next lookback. Settled inside the audit rather than after it, the renewal is leverage. Settled after, it is a separate bill.