Resolve Desk Queue Hygiene for Mid-Market Teams
What breaks without hygiene
- Renewal surprises — open P1 on account nobody flagged to CSM
- Duplicate effort — three cases for one outage
- SLA theater — P2 aging because queue sort is wrong
- Sales blind spots — AE pitches expansion while desk shows churn risk
- Agent garbage — MCP summarizes 400 stale cases nobody trusts
Product overview: Resolve service desk. Line choice: Momentum vs Resolve.
P1 / P2 definitions (write these down)
| Priority | Definition | Response target |
|---|---|---|
| P1 | Production down, security, billing blocking work | Same day — often hours |
| P2 | Major feature broken; workaround painful | 1–2 business days |
| P3 | Standard how-to, config, non-blocking bug | 3–5 business days |
| P4 | Feature request, nice-to-have | Backlog honest ETA |
Adjust targets to your team size — publish them inside Resolve so embed and humans align.
Monday desk ritual (30 minutes)
- P1/P2 sweep — any breach? escalate before standup
- Unowned cases — assign owner or merge duplicate
- No account link — match domain to account or create lead
- Stale >14 days — customer contact or resolve with note
- Renewal cross-check — open P1/P2 on accounts renewing in 90 days
Pair with revenue standup: weekly revenue standup.
Account linking rules
- Every case on an account — not free-text company name only
- Contact role — requester mapped to contact record
- Related opp — pre-sales cases link to open pipeline when relevant
- Mail on account — replies from org mail on case thread
- Support Embed handoffs — widget cases inherit same rules — Support Embed rollout
Sales ↔ desk handoff
| Event | Desk action | Sales visibility |
|---|---|---|
| Closed-won | Kickoff case template; CSM owner | AE sees case on account |
| Escalation during pilot | P1 until resolved; exec loop | AE + manager on thread |
| Renewal 90 days | Scrub open P2+ on account | CSM renewal plan in Workspace |
| Expansion signal | Tag case category; notify AM | Opp task from desk note |
Handoff checklist: sales to CS handoff.
Duplicate and merge discipline
Same outage, five cases — merge to one parent, link mail threads, single owner. Rules:
- Merge within 24h of duplicate detection
- Primary case keeps SLA clock — secondary resolved as duplicate
- Customer gets one thread — no parallel replies from two agents
MCP prompt (desk hygiene read)
List open P1 and P2 cases with no mail update in 7 days, grouped by account. Flag accounts with renewal date in next 90 days — read only.
Token and write policy: token governance, Salestrics MCP.
Monthly metrics review
| Metric | Why |
|---|---|
| Median time to first response | SLA honesty |
| Open cases per agent | Load balance |
| % cases with account link | Graph completeness target 100% |
| Reopen rate | False resolves |
| P1 on renewing accounts | Renewal risk — should trend down |
Anti-patterns
- Personal Gmail replies — breaks mail on account and MCP
- Priority inflation — everything P1 means nothing is P1
- Zombie queue — cases open months with no touch
- Desk in silo — no Monday cross-read with sales pipeline
- Skip Resolve tier — desk on external tool while CRM is Salestrics
90-day hygiene maturity
- Month 1 — owners and account links on all open cases
- Month 2 — SLA published; embed + desk same priority rules
- Month 3 — MCP Monday read standard; renewal cross-check automated in ritual
Renewal depth: renewal playbook and first renewal conversations.
Agent workload and queue caps
Hygiene fails when every agent owns 80 open cases. Mid-market teams should publish caps:
| Role | Open case cap | When exceeded |
|---|---|---|
| Tier 1 agent | 25 active | Manager redistributes or hires |
| Tier 2 / escalation | 10 active P1/P2 | Exec loop on breach risk |
| CSM (non-desk) | 5 owned cases max | Hand back to desk queue |
Caps are not headcount formulas — they are hygiene guardrails. When an agent exceeds cap, Monday ritual assigns overflow to backup owners or escalates hiring conversation with finance.
Category taxonomy (keep it small)
Too many case categories create miscategorization and useless reporting. Start with six:
- Technical — bugs, outages, integrations
- Billing — invoices, refunds, seat changes
- How-to — configuration, training, docs gaps
- Pre-sales — pricing, demo, evaluation support
- Renewal / expansion — contract, upsell, downgrade risk
- Other — catch-all reviewed monthly for new category candidates
RevOps maps categories to renewal dashboards once per quarter — not ad hoc per ticket. Align with agentic support handoff so sales sees the same labels on account timeline.
Worked example: Monday ritual in 30 minutes
A 40-person B2B company runs Resolve Team with three desk agents:
- Minutes 0–5 — P1/P2 sweep: two P2 breaches flagged; one reassigned, one escalated
- Minutes 5–15 — Unowned queue: four cases get owners; two duplicates merged
- Minutes 15–22 — Account link pass: three cases matched by domain to existing accounts
- Minutes 22–28 — Renewal cross-check: one account renewing in 60 days with open P2 — CSM pinged
- Minutes 28–30 — MCP read prompt queued for CS lead: stale P2 list for afternoon review
Output: queue honest before customer calls; no surprise renewal risk; agents start the week with clear ownership instead of re-triaging Friday leftovers.