Reviewing incident response commitments with a vendor

Review a vendor’s incident commitments: which events qualify, which functions are covered, when each clock starts, how updates arrive and what remedy applies. Read availability, support and recovery terms separately. A first-response target does not promise restoration; a monthly uptime figure does not describe the team’s fallback work.

Build a commitment sheet for the proposed plan

Identify the documents and versions incorporated into the proposed agreement. Record the product and plan, covered functions, measurement period, exclusions, incident route, severity rules, support hours, first-response target, recovery objective and claim procedure. Name someone who can raise an incident if the normal helpdesk is inaccessible.

Use the team’s critical paths to question each clause. If customers can send messages but agents cannot reply, does the availability definition count that as downtime? If an external messaging provider fails, is the connection covered?

Can a partial outage affecting Australian customers be raised and tracked? Seek written answers for the exact proposed setup.

Commitment sheet items to record

Compare published terms by what they measure

The Intercom SLA makes its termination and refund right the sole and exclusive remedy for failure to meet target availability.

Intercom: availability and support response

Intercom’s published agreement defines its Core Platform as Messenger and Inbox, using key actions such as replying to assess uptime. It sets a 99.8% monthly availability target for the Core Platform and Fin AI Agent, subject to exclusions including third-party systems and maintenance.

For a verified failure in two consecutive months, its stated remedy is a conditional termination and refund right for the affected service. Its Essential, Advanced and Expert support policy gives target first responses for urgent and non-urgent requests but expressly does not guarantee a specific response time. Check which required actions and services the proposed agreement covers.

Atlassian Jira Service Management: covered experiences

Atlassian’s cloud SLA lists 99.9% for Premium and 99.95% for Enterprise eligible products. Eligibility is limited to Cloud Products listed in Appendix A; the commitment does not include products that the customer has not provisioned.

Atlassian’s monitoring and logging infrastructure is the sole source of truth for determining whether it has met the service level commitment. To seek a service credit, a customer must submit a fully completed support ticket within 15 days after the end of the month in which the alleged failure occurred. Compare the claim process and eligibility requirements with the proposed agreement.

Zendesk: a defined disaster-recovery feature

Zendesk documents Enhanced Disaster Recovery with Enterprise Plus. For a qualifying disaster, it aims to restore normal Zendesk Support account operations within four hours of disaster declaration, except when all Availability Zones used by an account are affected.

Its recovery point objective targets one hour or less of data loss measured from disruption. These are objectives for a defined disaster-recovery feature, not a general incident response or integration uptime promise.

Published vendor terms: what each commitment measures

Make the operating decision

Mark each critical requirement as covered, excluded or unclear. Seek written clarification for an unclear clause and confirm the account’s actual contract and entitlement. Record the vendor contact route alongside the team’s fallback for requests, essential records and customer updates.

Make the operating decision