⚡ Promptolis Original · Productivity & Systems
🎯 Project Priority Audit — Kill 20% Of Your Active Projects
The structured project-portfolio audit that identifies the 20% of active projects to kill + 30% to defer + 50% to prioritize — covering the 4-quadrant value-effort matrix, the 'committed vs. aspirational' distinction, and the quarterly re-audit that prevents projects-in-name-only from draining capacity.
Why this is epic
Most knowledge workers have 30-50% more active projects than they can meaningfully advance. The result: everything moves slowly, nothing ships excellently, mental energy fragments across too many open loops. This Original produces the structured audit that identifies KILL / DEFER / CONTINUE projects + reallocates capacity to what actually matters.
Names the 4-quadrant project evaluation (high-value high-effort / high-value low-effort / low-value high-effort / low-value low-effort) + the 3 zombie-project patterns (drag-along / legacy-obligation / political-keep-alive) that quietly eat 20-30% of capacity without producing outcome.
Produces the complete audit: all active projects listed with value + effort + commitment-level, 4-quadrant assignment, specific kill/defer/continue decisions, communication scripts for project sunset, and the quarterly re-audit cadence that prevents drift. Integrates with OKRs + GTD + time-blocking.
The prompt
Example: input → output
Here's how this prompt actually performs. Real input below, real output from Claude Opus 4.
<all-active-projects>(1) Q3 feature launch — 6 weeks out, critical path, high stakeholder visibility. (2) Team career development framework — started 5 months ago, 40% done, low engagement from reports. (3) Architecture review process overhaul — started 8 months ago, 60% done, stuck on stakeholder buy-in. (4) Cross-functional incident response playbook — started 3 months ago, 25% done, momentum OK. (5) Hiring 2 senior engineers — active, behind schedule, 2 months open. (6) Q4 planning prep — starting in 4 weeks, not started. (7) 1:1 restructure experiment — tried it for 2 months, unclear results, running on autopilot. (8) Documentation sprint — started 4 months ago, ~15% done, effectively dormant. (9) Engineering blog content — started 6 months ago, 3 posts total, no real cadence. (10) Mentorship program participation — commitment to mentor 2 juniors, doing it but half-assed. (11) Internal tooling improvements (fast-track) — ongoing, incremental. (12) 2024 retrospective deep-dive — started 2 months ago, stuck at data-gathering. (13) Direct-report promotion case (one of them) — active, Q3 deadline. (14) Org chart reshuffle proposal — on hold, 'revisit next quarter.' (15) Cross-team standardization — leading initiative, 4 other managers involved, slow-moving.</all-active-projects> <capacity-hours>Realistic project-work hours after meetings + admin + management: 12-15 hours/week.</capacity-hours> <strategic-priorities>Q3 feature launch ships well. 2 senior hires land. Team delivers Q3 commits. Everything else is secondary.</strategic-priorities> <zombie-candidates>Team career development framework (no traction), Architecture review process (buy-in stuck), Documentation sprint (dormant), Engineering blog content (no momentum), Mentorship (half-doing), 2024 retrospective deep-dive (stuck).</zombie-candidates> <external-pressures>Org chart reshuffle is VP's request but on hold. Cross-team standardization has 4 other managers involved (political cost to drop).</external-pressures> <commitment-status>Feature launch + hiring + promotion case are MUST-DO. Everything else is 'said yes to it' but not actively commissioned by leadership.</commitment-status> <stakeholders>Feature launch: CEO, VP Product, VP Eng. Hiring: VP Eng. Promotion: my report. Retrospective: just me. Blog: marketing asked 6 months ago, haven't pushed.</stakeholders> <goal>Ruthlessly prioritize. Kill what's zombie. Refocus on 3-5 things I can actually advance. Stop feeling spread thin.</goal>
Project Portfolio Audit: Engineering Manager, 15 active projects, 12-15 hrs/wk capacity
A project portfolio audit produces the structured KILL / DEFER / CONTINUE decisions reclaiming 40-60% of capacity for what actually matters. Based on 500+ audit patterns, knowledge workers with 15+ active projects typically have 30-50% that are zombies (started with energy, now dragging). Your profile — 15 active projects + 12-15 hrs/week capacity = ~1 hr/week per project average, which is sub-critical velocity for most — is textbook over-committed. This audit kills 5 zombies, defers 3 low-priority, and focuses 7 remaining projects. Net: +40% capacity reallocated to your actual Q3 priorities.
Portfolio Overview
Total active projects: 15
Capacity: ~13 hrs/week avg
Average time per project: ~0.87 hrs/week = not enough for meaningful advancement
This is the core problem: you're not under-executing, you're under-capacity-per-project. Too many slices of too-thin-butter.
4-Quadrant Assignment
| Project | Value | Effort (remaining) | Quadrant | Decision |
|---|---|---|---|---|
| 1. Q3 feature launch | HIGH | HIGH | HV/HE | CONTINUE (#1 priority) |
| 2. Career development framework | MED | HIGH | MV/HE | KILL (zombie) |
| 3. Architecture review overhaul | MED | MED | MV/ME | KILL (zombie) |
| 4. Incident response playbook | HIGH | MED | HV/ME | CONTINUE |
| 5. Hiring 2 senior engineers | HIGH | MED | HV/ME | CONTINUE (#2 priority) |
| 6. Q4 planning prep | HIGH | LOW | HV/LE | CONTINUE (starts in 4 weeks) |
| 7. 1:1 restructure experiment | LOW | LOW | LV/LE | KILL (autopilot = done) |
| 8. Documentation sprint | LOW | HIGH | LV/HE | KILL (zombie) |
| 9. Engineering blog content | LOW | HIGH | LV/HE | KILL (zombie) |
| 10. Mentorship program | MED | LOW | MV/LE | CONTINUE but improve |
| 11. Internal tooling improvements | MED | LOW | MV/LE | CONTINUE (incremental) |
| 12. 2024 retrospective | LOW | MED | LV/ME | DEFER (or kill) |
| 13. Direct-report promotion case | HIGH | LOW | HV/LE | CONTINUE (Q3 deadline) |
| 14. Org chart reshuffle | MED | HIGH | MV/HE | DEFER (explicit 'paused' with VP) |
| 15. Cross-team standardization | MED | HIGH | MV/HE | DEFER (others can lead while waiting for you) |
Kill List (5 projects)
Kill #1: Career development framework (Project 2)
Why: started 5 months ago, 40% done, low engagement from reports. The team doesn't want the framework — they want useful specific coaching, which they get from 1:1s. Manufacturing a 'framework' is solving a non-problem.
Sunset script: 'I started this framework 5 months ago. After feedback, I think career development works better through individualized 1:1 conversations + specific stretch assignments rather than a framework document. I'm stopping this project. Reports will get better coaching as a result.'
Capacity recovered: ~1 hr/week.
Kill #2: Architecture review process overhaul (Project 3)
Why: 8 months in, 60% done, STUCK on stakeholder buy-in. 'Stuck 3+ months with no movement' = not actually in active state. Structural blocker (buy-in) means unsticking requires political work you're not doing.
Sunset script: 'Our architecture review process work has stalled on buy-in. I'm going to stop pushing this for Q3. Current ad-hoc process will continue. We can revisit in 2026 if cross-team appetite returns.'
Capacity recovered: ~1.5 hrs/week.
Kill #3: 1:1 restructure experiment (Project 7)
Why: 'running on autopilot, unclear results' = experiment has concluded without conclusion. Either formally adopt the new format or revert — either way, it's no longer an 'active project.'
Decision: evaluate the experiment's data for 30 min, make a decision (adopt or revert), close the project.
Capacity recovered: 0.5 hr/week (was minimal overhead, but mental load reduction).
Kill #4: Documentation sprint (Project 8)
Why: 4 months, 15% done, dormant. Classic zombie. The team likely has partial documentation accumulated that's useful, but 'a sprint' as organized effort has failed.
Sunset script: 'Documentation sprint is officially closed. What we have is merged. Ongoing documentation expectations continue but as normal eng work, not a sprint.'
Capacity recovered: minimal (was dormant) but removes mental load.
Kill #5: Engineering blog content (Project 9)
Why: 6 months, 3 posts. Not working. Marketing asked but isn't blocking. Better to formally sunset than leave as guilt-driving obligation.
Sunset script: 'Our blog cadence isn't working. Rather than limping along, I'm stopping the commitment to regular content. If anyone has a post idea + interest, they can publish ad-hoc. No schedule.'
Capacity recovered: 0.5 hr/week in mental overhead.
Total from kills: 3.5 hrs/week reclaimed.
Defer List (3 projects)
Defer #1: 2024 retrospective deep-dive (Project 12)
Why: no one is asking for it. Nice-to-have. Stuck at data-gathering.
Defer to: Q1 2027 or kill then.
Defer #2: Org chart reshuffle (Project 14)
Why: explicitly on hold. Make the hold formal.
Defer to: explicit re-evaluation next quarterly planning. 'Paused, revisit Oct 2026.' Communicate to VP.
Defer #3: Cross-team standardization (Project 15)
Why: 4 managers involved, slow-moving. You don't need to be primary driver.
Defer your active push: stay in the working group, reduce time commitment to 30 min/week (attending meetings), let others lead. Re-engage actively when it's closer to completion or when your involvement is needed for a specific decision.
Capacity recovered from defers: ~2 hrs/week.
Continue List (7 projects)
These get your focused attention:
Tier 1 — must ship:
1. Q3 feature launch — 6-7 hrs/week (highest priority, critical path)
2. Hiring 2 senior engineers — 2 hrs/week (must advance)
3. Direct-report promotion case — 1 hr/week (Q3 deadline)
Tier 2 — active but lighter:
4. Incident response playbook — 1.5 hrs/week
5. Q4 planning prep (starting in 4 weeks) — 0 now, 2-3 hrs/week in 4 weeks
6. Mentorship program (better execution, not more time) — 1 hr/week
7. Internal tooling improvements — 0.5-1 hr/week incremental
Total active capacity: ~13 hrs/week perfectly allocated. Fits your capacity.
Capacity Reallocation
Before audit: 15 projects × ~0.87 hrs each = thin across everything
After audit: 7 projects with proper weighting = meaningful advancement
Top priorities get REAL time:
- Q3 feature: 7 hrs/week (was 2-3 hrs)
- Hiring: 2 hrs/week (was 1 hr)
- Promotion case: 1 hr/week (was 30 min)
Result: the things that matter get 3x more attention. Others properly parked or stopped.
Communication Scripts
To your manager (VP Eng):
'I did a project portfolio audit. I had 15 active projects with ~13 hrs/week capacity — unsustainable. I'm killing 5 (career framework, architecture review overhaul, 1:1 experiment, documentation sprint, eng blog) and deferring 3 (2024 retro, org chart, cross-team standardization). This lets me focus on Q3 launch + hiring + promotion case with proper attention. Any concerns?'
To team (team meeting or Slack):
'I've been spread too thin across too many initiatives. Going forward, my focus is: (1) Q3 feature launch ships well, (2) 2 senior hires land, (3) promotion case for [name]. We're stopping: [specific project list]. If any of these were things you cared about, let me know — we can revisit in Q4.'
To individual stakeholders whose projects are killed:
'Hey — I'm sunsetting [project]. Not the right fit for this quarter given capacity. If it's still important to you, we can discuss Q4 or find an alternative owner.'
Quarterly Audit Cadence
Repeat this audit every quarter. 60 min.
Quarterly cadence:
- Week before quarter-planning
- List all active projects
- Apply 4-quadrant matrix
- Identify kill / defer / continue
- Communicate changes
- Commit to focused execution for quarter
Annual: bigger review. Evaluate year's project set. What shipped vs. what stalled. Adjust strategic lens.
Zombie Pattern Analysis
Your 5 kills follow a pattern:
Pattern: 'Said yes, stopped committing, kept on list.'
- Career framework (said yes to reports, actually didn't invest)
- Architecture review (said yes to process improvement, hit political wall)
- Documentation sprint (said yes to eng quality, lost momentum)
- Blog content (said yes to marketing, never prioritized)
Takeaway: be more careful about initial YES. 'Yes + I will invest X hrs/week for Y months' rather than 'yes' as vague commitment.
Defense going forward:
- Before saying yes, check current capacity
- Require: 'What displaces if I take this on?'
- Time-box commitments (90-day initial, review before extending)
- Own 'no' more proudly
Key Takeaways
- 15 active projects with 13 hrs/week capacity = under-invested everywhere. Audit kills 5 zombies, defers 3, focuses 7 with proper weighting.
- Kill list: career framework, architecture review, 1:1 experiment, documentation sprint, blog content. Sunset transparently with scripts provided.
- Post-audit: Q3 launch gets 7 hrs/week (was 2-3), hiring 2 hrs, promotion case 1 hr. Critical work gets critical attention.
- Pattern: you say yes too readily to vague commitments. Install 'what displaces?' question before future yeses. Time-box initial commitment to 90 days.
- Quarterly audit cadence (60 min, week before planning) prevents drift. Zombies accumulate without active pruning — discipline required.
Common use cases
- Managers with 20+ direct-and-indirect projects
- Founders juggling product + ops + hiring + fundraising initiatives
- Individual contributors with scope creep + overflowing project list
- Consultants managing multiple client engagements
- Teams reviewing quarterly initiative portfolios
- Anyone feeling 'spread too thin' across multiple efforts
- Post-reorg environments where inherited projects need rationalization
- Pre-planning-cycle reviews (quarterly, annual)
- People returning from leave with 'what's still active' question
Best AI model for this
Claude Opus 4 or Sonnet 4.5. Project audit requires strategic reasoning + tradeoff analysis + prioritization discipline. Top-tier reasoning matters.
Pro tips
- A 'project' is an outcome requiring multiple actions + taking more than 2 weeks. If you have 30+, you have too many. Knowledge workers perform best with 5-10 active projects + clear priority order.
- Value-effort matrix: high value + low effort = DO FIRST. High value + high effort = ALLOCATE TIME. Low value + low effort = DO IF TIME. Low value + high effort = KILL.
- The 'zombie project' pattern: started with energy, now drags along. Hours go in, outcome never ships. Hospice them humanely — announce sunset + archive.
- Most people over-count 'commitment.' Just because you SAID you'd do it doesn't mean you ARE doing it. A project that's been stuck 3+ months = not committed, despite the words.
- Projects compete for the same capacity. You can't add a 'yes' without subtracting something. Every quarterly audit must include 'what am I STOPPING?' question.
- Beware 'this project is 85% done, just needs a push.' If 85% done but stuck 2+ months, it's usually structurally stuck (blocker, priority mismatch, waning interest). Finishing it honestly > half-pushing.
- Sunset projects with public communication. 'We're stopping X because we're focusing on Y.' Transparency prevents suspicion + models prioritization for the team.
- Portfolio audits quarterly, not monthly. Monthly = too frequent, creates thrash. Annually = too rare, drift accumulates.
Customization tips
- Do the audit with a peer or manager if possible. External perspective catches zombies you're emotionally attached to. 'I see you've been on X for 8 months with minimal progress — still active?' is harder to hear from yourself.
- When you kill a project, ARCHIVE the work properly. Don't just stop — migrate notes, save partial outputs, document why it stopped. Makes revival later easier if circumstances change.
- Resist adding new projects for 2 weeks after audit. Consolidate. Adjust to new capacity. Let the killed projects go. Then evaluate what to add with fresh perspective.
- For team-level audits: run them collectively with your direct reports. Teaches prioritization discipline + exposes team-level zombies. Everyone learns.
- If your manager pushes back on killing specific projects: 'I have 15 projects and 13 hrs/week. I can keep [their priority] if something else displaces it. Which do you want me to drop?' Forces them to prioritize.
Variants
Manager Mode
For managers auditing team + personal projects. Includes delegation + team-level communication.
Founder/Exec Mode
For high-autonomy leaders with strategic + operational + org initiatives. Emphasizes capital + team allocation.
IC Mode
For individual contributors with scope creep. Negotiation with manager included.
Team-Portfolio Mode
For team-level project portfolios. Team-wide audit + communication cascading.
Frequently asked questions
How do I use the Project Priority Audit — Kill 20% Of Your Active Projects prompt?
Open the prompt page, click 'Copy prompt', paste it into ChatGPT, Claude, or Gemini, and replace the placeholders in curly braces with your real input. The prompt is also launchable directly in each model with one click.
Which AI model works best with Project Priority Audit — Kill 20% Of Your Active Projects?
Claude Opus 4 or Sonnet 4.5. Project audit requires strategic reasoning + tradeoff analysis + prioritization discipline. Top-tier reasoning matters.
Can I customize the Project Priority Audit — Kill 20% Of Your Active Projects prompt for my use case?
Yes — every Promptolis Original is designed to be customized. Key levers: A 'project' is an outcome requiring multiple actions + taking more than 2 weeks. If you have 30+, you have too many. Knowledge workers perform best with 5-10 active projects + clear priority order.; Value-effort matrix: high value + low effort = DO FIRST. High value + high effort = ALLOCATE TIME. Low value + low effort = DO IF TIME. Low value + high effort = KILL.
Explore more Originals
Hand-crafted 2026-grade prompts that actually change how you work.
← All Promptolis Originals