Why service levels matters 

A service level regime is essentially a contract within a contract. It sets out measurable performance standards the vendor must meet and the consequences of failing to meet those standards. When implemented properly, service levels serve several critical purposes:

• they define what ‘good performance’ actually looks like in practice;
• they reduce operational risk by establishing clear response, resolution and escalation processes;
• they support business continuity and disaster recovery planning; and
• they create an ongoing incentive for the vendor to meet (and ideally improve) performance over time.

Service levels are particularly important where the technology or services are business‑critical, the vendor has operational control (as is often the case with SaaS or other cloud services), or where downtime or a security incident would have serious financial, operational or reputational consequences.

Measure what matters (not necessarily what is easy)

One of the most common pitfalls in service level design is the selection of metrics that are easy to measure but commercially irrelevant. For example, fast ‘response times’ (i.e. times for the vendor’s support team to acknowledge a support ticket) are of little comfort if issues are not resolved within an acceptable timeframe.

Effective service levels focus on outcomes that matter to the customer’s business. Depending on the solution, this might include:

• availability and uptime for SaaS platforms;
• incident response and resolution times for managed services (including, where relevant, cybersecurity incident response and notification obligations);
• recovery time objectives (RTO) and recovery point objectives (RPO) for hosted or data‑intensive solutions; 
• performance and speed metrics that reflect the actual user experience; and/or
• other scope-specific service levels. 

Identifying the right service levels requires input from business users and technical stakeholders, not just lawyers. While careful legal drafting is essential to ensure enforceability, only those who rely on the system or services on a day‑to‑day basis can properly assess which performance standards are genuinely important.

Be precise in definitions

Service levels will only operate effectively where the methodology for calculating performance is clearly defined and agreed. Ambiguous drafting creates scope for dispute and undermines the effectiveness of the service level regime.

Customers should ensure that service levels clearly specify:

• the performance metrics being measured (including how concepts such as ‘availability’ are defined);
• the methodology and timeframe for measurement;
• any applicable exclusions (including scheduled maintenance during agreed ‘out of hours’ windows); and
• the criteria for determining incident severity or priority levels.

For availability metrics in particular, customers should carefully scrutinise carve‑outs. Excessive or broadly worded exclusions (e.g. excusing downtime arising from ‘emergency maintenance’ or ‘security incidents’) can significantly undermine the commercial value of the service level regime. Customers should resist carve-outs that excuse the vendor from performance obligations arising from events within the vendor’s reasonable control, including incidents caused by the vendor’s failure to maintain adequate security or resilience measures.

Make service levels objectively measurable

If performance cannot be objectively measured, service levels will not operate effectively. Automated monitoring tools, dashboards, alerts and regular performance reporting often provide greater accuracy and transparency than reliance on manual reporting.

Where the vendor is responsible for monitoring its own compliance, customers should consider:

• the level of detail required in performance reports;
• rights to audit or independently verify performance data; and
• whether the measurement and reporting obligations are proportionate to the value and risk profile of the services.

Overly complex service level regimes can increase cost and administrative burden without delivering better outcomes. The focus should remain on measuring the ultimate performance outcome, rather than every step in the service delivery process.

Use service credits as incentives, not penalties

Service credits are a common remedy for service level failures, but they must be carefully structured. 

From a legal perspective, service credits should be structured so that they are not out of proportion to the customer’s legitimate interests in performance, including by reference to a genuine pre‑estimate of likely loss, to avoid being characterised as unenforceable penalties.

From a commercial perspective, service credits must be sufficiently meaningful to incentivise performance. 

Key considerations include:

• scaling service credits to reflect the severity and duration of the failure;
• applying escalation mechanisms for repeated or systemic breaches; and
• avoiding ‘one‑size‑fits‑all’ models that fail to reflect actual business impact.

A minor but recurring failure may be just as disruptive as a single major incident. Sliding scales, tiered models or points‑based systems can help capture cumulative impact and discourage vendors from treating service credits as a cost of doing business.

Avoid service credits as the sole remedy

Vendors frequently seek to provide that service credits constitute the customer’s sole and exclusive remedy for service level failures. This is a high‑risk position for customers, especially where service credits are capped.

Sole and exclusive remedy regimes operate as an additional and often overlooked limitation on a vendor’s liability. In practice, such regimes can materially restrict the customer’s ability to recover loss, resulting in a lower recovery than would otherwise be available if no service credit regime applied.

Customers should, wherever practicable, preserve:

• all contractual and statutory rights and remedies, including the right to recover damages at law where losses exceed service credit thresholds; and
• express termination rights for material or repeated failures to meet service levels.

Where vendors resist this position, a common compromise is to allow customers to elect between service credits and damages, or to preserve termination rights (and the right to claim resultant damages) even where service credits are expressed to be the sole financial remedy. 

For business-critical services, customers should also consider negotiating step-in rights, which allow the customer to assume operational control of service delivery, or to engage a third party at the vendor’s cost, if the vendor persistently fails to meet service levels. Step-in rights provide a practical remedy that goes beyond financial compensation.

Plan for change and long‑term contracts

Technology evolves quickly. Static service levels in long‑term contracts risk becoming outdated or misaligned with evolving business needs. Best‑practice service level regimes therefore incorporate:

• mechanisms for regular review and adjustment, ensuring that performance standards remain relevant and commercially appropriate over time; 
• benchmarking rights, allowing the customer to compare the vendor’s service levels against prevailing market or industry standards and to require improvement where the vendor is materially below benchmark; and 
• continuous improvement obligations, requiring the vendor to proactively identify and implement improvements to service delivery over the contract term, and not merely to maintain the status quo.

Final thoughts 

Service levels are not intended to punish vendors. Rather, they are designed to protect customers and promote consistent, high‑quality performance. When carefully designed, service level regimes reduce uncertainty, support operational resilience and provide a clear framework for managing issues when they arise. Poorly designed service level regimes, by contrast, can leave customers with inadequate remedies, limited leverage and unexpected gaps in protection, often at precisely the moment those protections are needed most.

If you would like assistance reviewing or negotiating service level provisions in a technology contract, please contact us.