Service levels work only when they bite.

The four building blocks of a real SLA.

An SLA that actually protects the buyer has four building blocks. A defined metric with a precise formula. A documented baseline that the parties agreed before signing. A measurement and reporting cadence with named owner. A service credit consequence that creates real economic incentive for the seller to perform. Without all four, the SLA is a sentence in the contract rather than a working operational tool.

Most first draft SLAs include the metric and skip the rest. The metric appears as a target, with no defined formula, no baseline data, no measurement methodology, and no consequence for missing it. The seller's view of how the metric is calculated may differ from the buyer's view. The dispute that follows consumes governance capacity and produces no useful outcome.

The discipline of a working SLA starts with the formula. How is uptime measured. What counts as scheduled downtime. What counts as a measurement period. How are partial outages weighted. The formula should be specific enough that the same data produces the same number regardless of who calculates it. Without this specificity, every monthly report becomes a discussion of methodology.

The baseline matters because targets without baselines are meaningless. A 99 percent availability target sounds substantive, but if the buyer never sees historical performance data, the target may be set well below actual performance. The buyer's posture should be that baselines are documented in the contract using actual data from the twelve months before signing.

Metrics that actually matter.

The metrics that matter in a TSA SLA are operational, not generic. For IT services, the relevant metrics include system availability, incident response time, incident resolution time, change request lead time, and data delivery accuracy. Each has a precise definition. Each can be measured. Each ties to a business outcome the buyer cares about.

For finance services, the metrics include close completion time, journal accuracy, accounts payable cycle time, accounts receivable aging, and reporting delivery dates. For payroll, the metrics include payroll accuracy, processing time, garnishment compliance, and year end reporting timeliness. For procurement, the metrics include cycle time from purchase requisition to purchase order, invoice processing time, and three way match accuracy.

The buyer should choose three to five metrics per major service area, not fifteen. Too many metrics dilute the discipline. The metrics that matter most are the ones tied to operational impact on Newco. A payroll error affects employee paychecks. An ERP outage affects month end close. The metrics should focus on outcomes the buyer can describe to the leadership team in simple terms.

Vanity metrics should be removed. Metrics that are easy to hit but operationally irrelevant. Metrics that are difficult to measure and easy to dispute. Metrics that the seller cannot influence directly. Each of these wastes contract space and dilutes the focus on what matters. The disputes that arise from poorly chosen metrics are covered in TSA exit failure modes.

Service credits that create real bite.

Service credits are the economic mechanism that turns an SLA from a target into a commercial commitment. The seller misses the metric. The seller pays a service credit to the buyer. The credit reduces the invoice in the period when the miss occurred. Without service credits, the SLA is aspirational. With them, the SLA shapes seller behavior.

The credit structure matters. A tiered structure that scales with the severity of the miss creates appropriate incentive. A small miss might produce a 2 percent credit on the affected service. A larger miss produces 5 to 10 percent. A material miss that triggers operational disruption produces 15 percent or more. The structure should reflect the actual cost of the miss to the buyer.

Earned automatically is the discipline that matters. The credit should be calculated by the seller in the monthly report and applied automatically to the invoice. The buyer should not have to claim each credit. The buyer should not have to negotiate each credit. Automatic application is the contractual default the buyer should require. Manual claim processes produce credits that get waived because the discipline of claiming them is lower than the discipline of providing the service.

Cumulative credits matter too. Repeated misses on the same metric should produce escalating credits. A first miss at 2 percent, a second miss at 5 percent, a third miss at 10 percent. The escalator creates incentive for the seller to invest in fixing the underlying issue rather than absorbing the credit each month as a cost of doing business. The cumulative escalator is one of the most negotiable terms because sellers resist it but rarely walk away over it.

Measurement and reporting cadence.

The measurement cycle should be monthly for most metrics. The seller measures, computes, and reports each metric within ten business days of month end. The report includes the metric value, the calculation, the underlying data, the comparison against target, and any earned service credits. The buyer reviews the report against its own monitoring and either accepts the result or raises a dispute through the change control process.

The buyer should also maintain independent measurement where possible. For metrics that can be measured from Newco's side, the buyer should measure them. Discrepancies between the seller's measurement and the buyer's measurement get resolved through the data, not through assertion. Without independent measurement, the buyer is dependent on the seller's view of performance.

The monthly SLA review is part of the standing governance rhythm. The performance against SLA is on the agenda at the workstream meeting and at the program governance committee. Repeated misses trigger escalation to the joint governance committee. Each level of review creates discipline. The seller knows that a single miss surfaces immediately, not at the next quarterly review.

The annual review of SLA structure is the moment to revisit. Targets that turn out to be too easy or too tight can be adjusted. Metrics that turn out to be irrelevant can be removed. New metrics for services that have evolved can be added. The annual review is most effective when both sides arrive with data, not preferences. The full governance pattern that supports SLA enforcement is covered in TSA exit governance best practices.

When to escalate beyond credits.

Service credits are not always enough. When the seller misses material metrics consistently, the buyer should escalate beyond credits. The contract should give the buyer rights to invoke remediation plans, to escalate to joint governance, and in serious cases to terminate the affected service. The escalation ladder is the structural commitment that backs the SLA.

A remediation plan is a structured response when SLA performance is consistently below target. The seller presents the root cause analysis, the remediation steps, the timeline, and the new performance expectations. The plan is reviewed at joint governance. Failure to execute the remediation triggers the next level of escalation. The mechanism prevents repeated misses from becoming the new normal.

Termination for service failure is the strongest remedy. When the seller's performance is so weak that the buyer cannot reasonably operate, the buyer should have the right to terminate the affected service. The buyer then either insources the service or transitions it to a third party. The termination right is rarely exercised, but the existence of the right shapes seller behavior in the cases that matter most.

The buyer's posture throughout should be commercial and direct. The contract gives the buyer specific rights. Exercising those rights is not adversarial, it is the operation of the agreement. The seller's deal team will sometimes characterize SLA enforcement as a relationship issue. The buyer should not accept that framing. SLA enforcement is contract enforcement, and the contract is the basis of the relationship.

More on TSA negotiation.