Scored automation backlog
A ranked backlog across Sales, Support, Engineering, and Finance. Adoption risk is the primary axis and is weighted heaviest — a high-impact tool nobody adopts is correctly ranked below a modest tool that sticks.
| Department · Workflow | Status | |||||
|---|---|---|---|---|---|---|
SalesBuilt Contract / RFP intake processor Extract structured terms from inbound venue contracts & RFPs; route low-confidence or PII fields to human review before they reach the CRM. Live in this console — see the Contract Processor tab. | 14 | 8 | 8 | 3 | 75 | Built |
SupportBuilt Reply / macro suggester Draft a first-pass reply to inbound support tickets from the knowledge base; agent edits and sends. Never auto-sends. | 22 | 7 | 9 | 4 | 69 | Built |
Engineering PR → release-note summarizer Summarize merged pull requests into customer-facing release notes; eng lead approves before publish. | 5 | 5 | 8 | 4 | 61 | Backlog |
FinanceBuilt Invoice GL-coding assistant Suggest general-ledger codes for inbound AP invoices; finance confirms. Deployed but adoption is soft. | 9 | 6 | 7 | 5 | 57 | Built |
Support Knowledge-base gap detector Cluster unanswered / deflected tickets to surface missing KB articles for the content team. | 7 | 6 | 6 | 5 | 55 | Backlog |
Finance Revenue-recognition anomaly flagger Flag contracts with non-standard rev-rec terms for finance review. High value, but feasibility and trust are unproven. High impact, but adoption risk pushes it below safer bets. | 8 | 8 | 5 | 6 | 54 | Backlog |
Engineering On-call incident triage drafter Draft an incident summary and suggested severity from logs for the on-call engineer to confirm. | 6 | 6 | 5 | 7 | 43 | Backlog |
Sales Proposal first-draft generator Generate a first-draft sales proposal from CRM opportunity data. High apparent value, but reps did not adopt it. KILLED — see Governance gate. 92% accuracy, 11% adoption. | 6 | 7 | 6 | 9 | 38 | Killed |
Opportunities come from a standing intake with each department lead, not a top-down list. In a short quarterly working session I capture the workflow, then we score impact and feasibility together — and the team that would actually use the tool owns the adoption-risk number, since they are the ones who decide whether it sticks. Scores are re-checked as tools ship: a built tool's real adoption (see the dashboard) feeds back into the board.
priority = ( impact·0.3 + feasibility·0.2 + (10 − adoptionRisk)·0.5 ) × 10Adoption fit (10 − adoption risk) carries the heaviest weight (0.5). Impact and feasibility together carry the other half. The score is recomputed live from the columns above, never stored.
Contract / RFP intake processor is the single highest-impact backlog item (impact 8), yet it ranks #1 of 7 active items — its adoption risk (3) pushes it below safer, stickier bets. The killed “Proposal first-draft generator” (impact 7) ranks last of all: high value means nothing if no one uses it.