Feature Backlog (with Conflicting Priorities)
Requirements elicitation artefact. A backlog of candidate CRM features. The priorities below are deliberately conflicting — each stakeholder has rated the same features differently. This file is designed for prioritisation exercises: MoSCoW, RICE, ICE, and WSJF all work against it. Do not treat the ratings as agreed; they are individual stakeholder positions to be reconciled.
1. How to read this backlog
Each feature carries:
- Requester — who asked for it (and therefore who will defend it).
- Business value — the analyst’s rough rating (H/M/L).
- Effort — a T-shirt size for delivery rough-order.
- MoSCoW as rated by each stakeholder — the source of the conflict.
The MoSCoW columns are: Must, Should, Could, Won’t (this time). Where stakeholders disagree, the row is marked ⚠️ in the Conflict column — these are the rows worth arguing over in class.
| Code | Meaning |
|---|---|
| CRM | Customer Relations Manager (Priya Nair) |
| SUP | Support Lead (Samantha Wong) |
| SAL | Sales |
| CFO | Chief Financial Officer (Aisha Rahman) |
2. The backlog
| ID | Feature | Requester | Value | Effort | CRM | SUP | SAL | CFO | Conflict |
|---|---|---|---|---|---|---|---|---|---|
| F01 | Single customer record (golden record, dedup of contacts) | CRM | H | L | Must | Must | Must | Must | — |
| F02 | Full audit trail (who/when for every field change) | CRM | H | M | Must | Could | Won’t | Must | ⚠️ |
| F03 | Support context screen (contract, products, recent tickets on call open) | SUP | H | M | Should | Must | Could | Should | ⚠️ |
| F04 | Helpdesk integration (read tickets into CRM; do not replace engine) | SUP | H | M | Must | Must | Should | Should | — |
| F05 | At-risk customer flag (low satisfaction + ticket frequency) | SUP | H | S | Should | Must | Must | Could | ⚠️ |
| F06 | Shared sales pipeline (entered once, visible to manager) | SAL | H | M | Must | Could | Must | Should | ⚠️ |
| F07 | Support-to-sales alerts (new high-priority ticket / satisfaction drop) | SAL | H | S | Should | Should | Must | Could | ⚠️ |
| F08 | Renewal workflow (reminders to the account owner, not just Customer Relations) | SAL | H | M | Must | Should | Must | Must | — |
| F09 | Mobile app (pipeline + alerts on the road) | SAL | M | L | Could | Won’t | Must | Won’t | ⚠️ |
| F10 | Xero two-way sync (contract value ↔︎ invoiced value) | CFO | H | M | Must | Could | Should | Must | ⚠️ |
| F11 | Board dashboard (active clients, at-risk renewals, revenue by product) | CRM | M | S | Must | Could | Should | Should | ⚠️ |
| F12 | AI chatbot for customer self-service ( deflect tier-1 queries ) | SAL | M | XL | Could | Won’t | Must | Won’t | ⚠️ |
| F13 | Marketing campaign module (email blasts, lead scoring) | SAL | L | L | Won’t | Won’t | Must | Won’t | ⚠️ |
| F14 | Custom report builder (ad-hoc queries for Finance/Audit) | CFO | M | M | Should | Could | Could | Must | ⚠️ |
| F15 | Single sign-on (SSO) with Microsoft 365 | CRM | M | S | Should | Should | Should | Must | — |
| F16 | Data migration tooling (spreadsheet → golden record, with dedup) | CRM | H | L | Must | Should | Should | Should | — |
| F17 | Role-based access control (field-level, e.g. contract value visibility) | CFO | H | M | Should | Should | Could | Must | ⚠️ |
| F18 | Australian data residency option (hosting in AU) | CFO | H | S | Must | Could | Could | Must | — |
| F19 | WhatsApp / SMS customer comms channel | SAL | L | M | Won’t | Could | Must | Won’t | ⚠️ |
| F20 | Automated CSAT survey after ticket closure | SUP | M | S | Should | Must | Could | Could | ⚠️ |
3. The conflicts worth teaching with
F09 (Mobile app) and F12 (AI chatbot) are both rated Must by Sales but Won’t by the CFO. Sales argues neither revenue nor retention is possible if reps and customers cannot reach the system on their terms; the CFO argues both are high-effort, high-recurring-cost features with unproven return. This is the classic value-vs-cost tension and a clean target for a RICE or WSJF exercise.
F02 (Audit trail), F06 (Shared pipeline), and F17 (Role-based access) collide on ownership. Sales rates shared pipeline Must but rates the audit trail Won’t — they want visibility without surveillance. The CFO rates the audit trail Must (defensibility) but Sales rates it Won’t (overhead). The CRM manager sits in the middle. This cluster is a good fit for a stakeholder-mapping + negotiation exercise.
F03 (Context screen) and F04 (Helpdesk integration) are Must for Support, but Sales leans toward a single unified UI that implicitly deepens the change. Samantha Wong has stated (see stakeholder needs, N4) she does not want the helpdesk engine replaced. Resolving F03/F04 against F12 (AI chatbot, which presumes a different support model) is where the architecture and the politics meet.
4. Conflicts summarised
| Cluster | Features at odds | Underlying tension |
|---|---|---|
| A | F09, F12 vs. CFO | Feature temptation vs. total cost of ownership |
| B | F02, F06, F17 | Data ownership & surveillance vs. shared visibility |
| C | F03, F04 vs. F12 | Integrate the existing engine vs. replace with AI-first model |
| D | F08 vs. F10 | Renewal catch-rate (revenue) vs. billing accuracy (defensibility) — both “Must”, competing for the same effort budget |
| E | F11 vs. F14 | Dashboards (CRM) vs. auditable custom reports (CFO) — different definitions of “good reporting” |
5. Prioritisation exercise prompts
- MoSCoW. Force the class to converge the four stakeholder columns into a single agreed MoSCoW rating per feature. Which conflicts resolve easily? Which need a trade-off?
- RICE. Score Reach, Impact, Confidence, Effort for F09, F12, F13, F19 (the “temptation” features). Does RICE rescue any of them from CFO rejection?
- WSJF / Cost-of-delay. Sequence F08 (renewals) and F10 (Xero sync) — which delivers more protected revenue per week of effort?
- Stakeholder mapping. For conflict cluster B, map each stakeholder’s power/interest and propose a negotiation outcome.
There is deliberately no “correct” prioritisation here. The artefact tests whether students can surface and negotiate trade-offs, not whether they can guess the answer the lecturer has in mind. Encourage explicit, defended trade-off logs.