sla-definer
Agent: VP Product
L1 product leader responsible for opportunity framing, MVP definition, PRD assembly, go-live coordination, and post-launch monitoring. Owns the product strategy and roadmap from idea through GA.
Department ethos: ideal-product.md
Skill Description
Defines the service-level agreements that the product will commit to for customers.
When to Use
- When a product is approaching GA and enterprise buyers require contractual SLA commitments
- When entering a new market segment (e.g., healthcare, financial services) with regulatory SLA expectations
- When post-incident reviews reveal that existing SLAs are misaligned with actual system capabilities
- When Sales escalations indicate that missing or vague SLA terms are stalling deal closure
Workflow
- Audit current capabilities: Gather historical data on uptime, p50/p95/p99 latency, incident frequency, MTTR, and support response times from engineering and ops. Deliverable: system performance baseline report.
- Benchmark against market: Research SLA terms offered by direct competitors and adjacent products in the same buyer segment. Note where the market floor and ceiling sit for each metric. Deliverable: competitive SLA benchmark.
- Define SLA metrics: Select the specific metrics to include — availability percentage, response-time percentiles, support-tier response windows, RTO/RPO for disaster recovery, and data-processing throughput guarantees. Deliverable: SLA metric definitions.
- Set target thresholds: Propose numeric targets for each metric that are achievable with current architecture and headroom for growth. Stress-test targets against worst-case scenarios from the incident history. Deliverable: proposed SLA targets with confidence intervals.
- Design remediation and credit policies: Define what happens when an SLA is breached — service credits, escalation paths, root-cause disclosure timelines, and customer notification protocols. Deliverable: SLA remediation policy.
- Review with Engineering and Legal: Validate that Engineering can instrument, measure, and alert on every SLA metric. Confirm that Legal approves the contractual language and liability exposure. Deliverable: engineering feasibility sign-off and legal review.
- Publish the SLA document: Finalize the SLA in both customer-facing and internal versions. The customer-facing version goes into contract templates; the internal version includes monitoring runbooks and escalation procedures. Deliverable: published SLA document (external + internal).
Anti-Patterns
- Aspirational SLAs: Committing to targets the system cannot currently meet, hoping infrastructure improvements will close the gap before a breach. Why: Breached SLAs erode customer trust and trigger financial credits that compound quickly.
- Copy-paste SLAs: Adopting a competitor's SLA verbatim without validating against your own architecture and ops maturity. Why: Your system characteristics differ; borrowed targets may be unachievable or too conservative.
- Metric-free SLAs: Using vague language like "high availability" or "best-effort response" without numeric commitments. Why: Ambiguous SLAs create disputes during incidents and undermine the credibility of the product in enterprise procurement.
- SLAs without monitoring: Publishing SLA targets without instrumented dashboards and alerting to detect breaches proactively. Why: If you cannot measure it, you cannot manage it, and customers will discover breaches before you do.
Output
On success: A published SLA document (customer-facing and internal) containing metric definitions, numeric targets, measurement methodology, remediation and credit policies, and monitoring runbook references — approved by Engineering, Legal, and Product.
On failure: Report which metrics lack sufficient historical data, what architectural gaps prevent commitment, the best available interim SLA (e.g., best-effort with transparency commitments), and a roadmap to close gaps for a full SLA in a defined timeframe.
Related Skills
business-model-sketcher — sibling skill under the same agent — combine with business-model-sketcher for end-to-end coverage
competitive-response-monitor — sibling skill under the same agent — combine with competitive-response-monitor for end-to-end coverage
goal-framer — sibling skill under the same agent — combine with goal-framer for end-to-end coverage
1---2name: sla-definer3description: This skill defines the service-level agreements that the product will commit to for customers, covering uptime, latency, support response times, and remediation policies. Use when preparing a product for GA or enterprise sales that require contractual SLAs. Also consider when renegotiating existing SLAs after architecture changes or incident post-mortems. Suggest when deal-desk escalations cite missing or unclear SLA language as a blocker.4---56# sla-definer78## Agent: VP Product9L1 product leader responsible for opportunity framing, MVP definition, PRD assembly, go-live coordination, and post-launch monitoring. Owns the product strategy and roadmap from idea through GA.1011Department ethos: [ideal-product.md](../../../../departments/product/ideal-product.md)1213## Skill Description14Defines the service-level agreements that the product will commit to for customers.1516## When to Use17- When a product is approaching GA and enterprise buyers require contractual SLA commitments18- When entering a new market segment (e.g., healthcare, financial services) with regulatory SLA expectations19- When post-incident reviews reveal that existing SLAs are misaligned with actual system capabilities20- When Sales escalations indicate that missing or vague SLA terms are stalling deal closure2122## Workflow231. **Audit current capabilities**: Gather historical data on uptime, p50/p95/p99 latency, incident frequency, MTTR, and support response times from engineering and ops. Deliverable: system performance baseline report.242. **Benchmark against market**: Research SLA terms offered by direct competitors and adjacent products in the same buyer segment. Note where the market floor and ceiling sit for each metric. Deliverable: competitive SLA benchmark.253. **Define SLA metrics**: Select the specific metrics to include — availability percentage, response-time percentiles, support-tier response windows, RTO/RPO for disaster recovery, and data-processing throughput guarantees. Deliverable: SLA metric definitions.264. **Set target thresholds**: Propose numeric targets for each metric that are achievable with current architecture and headroom for growth. Stress-test targets against worst-case scenarios from the incident history. Deliverable: proposed SLA targets with confidence intervals.275. **Design remediation and credit policies**: Define what happens when an SLA is breached — service credits, escalation paths, root-cause disclosure timelines, and customer notification protocols. Deliverable: SLA remediation policy.286. **Review with Engineering and Legal**: Validate that Engineering can instrument, measure, and alert on every SLA metric. Confirm that Legal approves the contractual language and liability exposure. Deliverable: engineering feasibility sign-off and legal review.297. **Publish the SLA document**: Finalize the SLA in both customer-facing and internal versions. The customer-facing version goes into contract templates; the internal version includes monitoring runbooks and escalation procedures. Deliverable: published SLA document (external + internal).3031## Anti-Patterns32- **Aspirational SLAs**: Committing to targets the system cannot currently meet, hoping infrastructure improvements will close the gap before a breach. *Why*: Breached SLAs erode customer trust and trigger financial credits that compound quickly.33- **Copy-paste SLAs**: Adopting a competitor's SLA verbatim without validating against your own architecture and ops maturity. *Why*: Your system characteristics differ; borrowed targets may be unachievable or too conservative.34- **Metric-free SLAs**: Using vague language like "high availability" or "best-effort response" without numeric commitments. *Why*: Ambiguous SLAs create disputes during incidents and undermine the credibility of the product in enterprise procurement.35- **SLAs without monitoring**: Publishing SLA targets without instrumented dashboards and alerting to detect breaches proactively. *Why*: If you cannot measure it, you cannot manage it, and customers will discover breaches before you do.3637## Output38**On success**: A published SLA document (customer-facing and internal) containing metric definitions, numeric targets, measurement methodology, remediation and credit policies, and monitoring runbook references — approved by Engineering, Legal, and Product.39**On failure**: Report which metrics lack sufficient historical data, what architectural gaps prevent commitment, the best available interim SLA (e.g., best-effort with transparency commitments), and a roadmap to close gaps for a full SLA in a defined timeframe.4041## Related Skills42- [`business-model-sketcher`](../business-model-sketcher/SKILL.md) — sibling skill under the same agent — combine with business-model-sketcher for end-to-end coverage43- [`competitive-response-monitor`](../competitive-response-monitor/SKILL.md) — sibling skill under the same agent — combine with competitive-response-monitor for end-to-end coverage44- [`goal-framer`](../goal-framer/SKILL.md) — sibling skill under the same agent — combine with goal-framer for end-to-end coverage