How Distil Processes Feedback
Understand the journey feedback takes from collection to shipped feature.
The Feedback Lifecycle
Every piece of feedback in Distil follows this path:
Collection → Evaluation → Decision → Execution → Closure
Let's break down each stage.
Stage 1: Collection
What happens: Feedback enters Distil from various sources.
Sources:
- Manual entry (you create cards)
- Email forwarding
- API integrations
- Slack (coming soon)
Outcome: Card is created in Needs Signal status.
Who's responsible: Anyone can collect feedback (PMs, CS, sales, support).
Time: Ongoing - feedback arrives whenever it arrives.
Learn more about collecting feedback
Stage 2: Evaluation
What happens: You review the card and ask "Is this signal or noise?"
Key questions:
- Is this a real pattern or a one-off?
- Have we heard this before?
- Does it align with our strategy?
- What problem are they actually trying to solve?
Actions:
- Add signal notes to capture context
- Tag the card for categorization
- Search for duplicates
- Investigate (talk to customers, check analytics)
Outcome: You decide whether to accept or archive.
Who's responsible: Product managers (occasionally with input from eng/design).
Time: Weekly review sessions (30-60 min).
Learn more about signal vs. noise
Stage 3: Decision (Acceptance)
What happens: You commit to building this by accepting the card.
What acceptance means:
- This is validated as worth building
- It moves to your roadmap (Accepted section)
- You've documented why it matters (rationale)
Requirements:
- Clear evidence this is signal
- Alignment with product strategy
- Documented rationale (especially with Governance Pack)
Outcome: Card moves to Accepted status.
Who's responsible: Product managers (with governance approval if needed).
Time: During evaluation - happens as you review cards.
Stage 4: Scoping
What happens: You plan how to build it.
Actions:
- Write implementation notes
- Break down into sub-tasks (if complex)
- Estimate effort (optional - can do this in Jira/Linear)
- Prioritize against other accepted work
Outcome: Card is ready to push to dev tools.
Who's responsible: Product managers (with engineering input).
Time: During sprint planning or roadmap sessions.
Stage 5: Execution
What happens: Engineering builds it.
Actions:
- Push card to Jira or Linear
- Card status becomes Ready
- Eng picks up the issue and builds
- Work happens in Jira/Linear, not Distil
Outcome: Feature is shipped.
Who's responsible: Engineering team.
Time: Depends on complexity and sprint schedules.
Learn more about pushing to Jira
Stage 6: Closure
What happens: You close the loop with requesters.
Actions:
- Mark card as shipped in Distil (optional)
- Update public roadmap if using Visibility Pack
- Email customers who requested it
- Add link to release notes or changelog
Outcome: Requesters know their feedback was heard and acted on.
Who's responsible: Product managers or customer success.
Time: When feature ships.
The Role of Status
Status tracks where feedback is in this lifecycle:
| Status | Stage | Meaning |
|---|---|---|
| Needs Signal | Evaluation | Unprocessed, needs review |
| Accepted | Decision/Scoping | Validated, on roadmap |
| Ready | Execution | Pushed to dev tools, being built |
| (Archived) | Rejected | Not worth building (noise) |
Decision Points
Accept or Archive?
After evaluation, you have two choices:
Accept if:
- Clear pattern (multiple requests)
- Strategic fit
- Validated problem
- Feasible to build
Archive if:
- One-off request with no supporting evidence
- Misaligned with product vision
- Solving the wrong problem
- Too vague or not actionable
Neither (leave in Needs Signal) if:
- Need more signal before deciding
- Waiting to see if it becomes a pattern
- Investigating feasibility
Push to Jira/Linear When?
Not all accepted cards should be pushed immediately.
Push when:
- Ready to build (this sprint or next)
- Scoped and prioritized
- Eng has capacity
Don't push:
- Accepted but not yet prioritized
- Waiting on dependencies
- Low priority backlog items
Result: Your Jira/Linear backlog stays lean and focused.
Time in Each Stage
Typical durations:
- Collection → Evaluation: 1-7 days (depends on review cadence)
- Evaluation → Decision: Minutes to weeks (depends on clarity of signal)
- Decision → Execution start: Days to months (depends on prioritization)
- Execution: Days to months (depends on complexity)
- Total cycle time: 2 weeks to 6+ months
Healthy targets:
- Evaluation time:
<7days for 90% of cards - Time in Needs Signal:
<30days average - Acceptance → Push:
<90days (don't let roadmap become a graveyard)
Bottlenecks to Avoid
Evaluation Bottleneck
Problem: Cards pile up in Needs Signal because you don't review regularly.
Solution: Weekly 30-min review sessions. Process 10-15 cards per session.
Acceptance Bottleneck
Problem: Everything stays in Needs Signal because you never accept anything.
Solution: Lower your bar slightly. Accepted doesn't mean "build tomorrow" - it means "validated as worth doing eventually."
Push Bottleneck
Problem: 200 cards in Accepted, engineering backlog is empty.
Solution: Push more aggressively during sprint planning. Accepted is for validation, not storage.
Feedback Collection Bottleneck
Problem: <5 cards/week coming in - not hearing from users.
Solution: Make feedback easier to give (email forwarding, in-app widget, proactive outreach).
Feedback Velocity Metrics
Track these to understand your process:
Collection rate: New cards per week (target: 20-50 for most products)
Evaluation rate: Cards processed per week (should match collection rate)
Accept rate: % of evaluated cards accepted (target: 20-40%)
Cycle time: Days from collection to shipped (varies widely, track p50 and p90)
Backlog size: Cards in Needs Signal (target: <75)
Processing at Scale
Solo PM (<100 cards/month)
- Review Needs Signal once per week (30 min)
- Accept/archive during review
- Push to Jira bi-weekly during sprint planning
Sustainable and manageable.
Small Team (100-300 cards/month)
- Distribute collection (CS, sales, support create cards)
- PM reviews and evaluates
- Weekly triage meeting (60 min)
- Bulk push during sprint planning
Requires discipline but doable.
Large Team (300+ cards/month)
- API integration for auto-collection
- Dedicated PM time for triage (daily 30 min)
- Team members pre-tag and add context
- PM accepts/archives, team executes
- Automation for duplicates and noise filtering
Requires tools and process.
The Role of Automation
What Distil automates:
- Duplicate detection (suggests similar cards)
- Tag suggestions based on content
- Status transitions (e.g., push → Ready automatically)
- Changelog generation (Visibility Pack)
What you still do manually:
- Evaluate signal vs. noise (requires judgment)
- Write acceptance rationale (requires context)
- Prioritize (requires strategy)
- Decide when to push (requires planning)
Philosophy: Automate mechanics, preserve judgment.
Collaboration in Processing
PM-Led
Model: PM does all evaluation and acceptance.
Pros: Consistent decisions, clear ownership Cons: PM is a bottleneck
Best for: Solo PM, small teams
Team-Assisted
Model: Team members add context and signal notes, PM decides.
Pros: Distributed collection, PM curates Cons: Requires coordination
Best for: Multi-PM teams
Fully Collaborative
Model: Multiple PMs evaluate and accept in their domains.
Pros: Scales well, domain expertise Cons: Requires clear ownership boundaries
Best for: Large product organizations
Learn more about team collaboration
Governance and Compliance
With Governance Pack enabled, the processing flow includes:
- Enforced rationale: Can't accept without documentation
- Audit trails: Every decision is logged
- Card locking: Prevent changes to accepted decisions
- Enhanced context: Jira/Linear issues show full rationale
Learn more about Governance Pack
Quality Control
Signs Your Process Is Healthy
✅ Needs Signal backlog stays <75 cards
✅ Cards don't sit unprocessed for >30 days
✅ Accept rate is 20-40% (not too high, not too low)
✅ Accepted cards get pushed within 90 days
✅ You can explain why any card was accepted or rejected
Signs Your Process Needs Work
❌ Needs Signal has 200+ cards
❌ Cards sit for months without review
❌ Accept rate is <10% (too conservative) or >70% (rubber-stamping)
❌ Accepted section has 100+ cards that never get pushed
❌ Can't remember why you accepted old cards
Next Steps
- Signal vs. Noise - Master evaluation
- What 'Accepting' Really Means - Decision philosophy
- Collecting Feedback - Improve collection
- Reviewing Cards - Efficient workflows