ACUTE PMC
Continue, Stabilize, Restructure, Stop, or Rebuild: A Decision Table for Troubled Technology Projects
COMMERCIAL DECISION THIS ENABLES: A sponsor can force a gates-first continue / stabilize / restructure / stop / rebuild call without treating a score or the incumbent SI as the decision authority.
—-
TL;DR
When a technology initiative is uncertain or failing, leadership needs a recommendation rule—not another status deck. This page defines five decision postures and the evidence criteria that support each call.
Hard gates are PASS/FAIL prerequisites. They are evaluated and displayed before scores. A failed gate stays visible and may eliminate postures even if later scores look strong.
- Scores are educational decision-support. They refine judgment among postures still allowed after gates.
- Scores do not qualify a buyer or select Project Clarity, Critical Project Rescue, or Critical Project Recovery.
- An arithmetic total cannot override a failed gate.
Why this decision is hard
Common failure modes:
Sunk-cost bias — We’ve spent too much to stop.”
- Watermelon status — green reporting / red reality.
- Vendor conflict of interest — the implementer authors “continue.”
- Conflating severity with ownership — redness alone does not decide who must lead execution.
Stop vs salvage vs rebuild is a sponsor decision, not an implementer sales path. Independent recovery leadership is not the same as an SI takeover.
Five outcomes — definitions (not service SKUs)
| Outcome | Meaning (educational) | Typical next ownership posture |
|—-|—-|—-|
| Continue under control | Business case still valid; delivery-evidence confidence high; decision latency acceptable | Client executes with existing leadership |
| Stabilize | Control/governance inadequate; capable receiving leadership exists | Bounded operating-system reset; client retains execution |
| Restructure | Scope/governance/vendor model wrong; path exists after redesign | Client-led redesign; may need specialist referral |
| Stop | Remaining value < cost-to-complete; or authority/evidence gap makes recovery unsafe | Controlled closure; preserve evidence/obligations |
| Rebuild or refer specialist | Technical foundation or regulated domain requires rebuild / specialist SI / legal-clinical authority Acute does not claim | Referral or separately scoped rebuild |
These postures describe **decisions**. Mapping any posture to an Acute intervention still requires human qualification: authority, sponsor, evidence, access, receiving leadership, capacity, and execution-ownership fit.
Hard decision gates (prerequisites — not scored)
Binary PASS/FAIL. Evaluated **before** any arithmetic. Display as a checklist **above** any scores. Never hide a FAIL behind a green composite. Gates never auto-route to Clarity / Rescue / Recovery.
| Gate ID | Prerequisite (educational) | If FAIL — constrained postures (examples) |
|—-|—-|—-|
| G1 | Named decision authority exists (sponsor/deputy who can bind) | Stop / pause for authority; Continue-under-control and Stabilize typically unavailable |
| G2 | Minimum decision-grade evidence accessible (or refusal is explicit) | Stop / pause; no honest Continue; Stabilize only after access path |
| G3 | **Economic/obligation basis is understood:** remaining value vs remaining cost is decision-grade; if remaining value is clearly below remaining cost **and** no overriding legal, safety, contractual, transition, or strategic obligation requires continuation, Continue/Rebuild optimism is constrained | **Stop / controlled wind-down must be explicitly evaluated**; Continue and Rebuild-optimism postures constrained when value≪cost with no overriding obligation |
| G4 | Legal/regulatory/safety showstopper absent (or owned by competent authority) | Stop / refer specialist; Acute does not claim regulated rebuild authority |
| G5 | A receiving owner for any next posture can be named (even if “sponsor for stop”) | Block Stabilize / Lead-Recovery-style postures that need a receiver |
**Display rule:** Show gate results as PASS/FAIL above any scores. A failed gate stays visible and may rule out Continue or Stabilize even if other scores are high.
Scored evidence dimensions (1–5, educational — subordinate to gates)
Score only **after** gates are recorded. Dimensions support judgment among **still-allowed** postures. They do **not** override a failed gate.
| Dim | Scored dimension | Role |
|—-|—-|—-|
| D1 | Business-case residual confidence | Fine-grained value vs cost (beyond G3 binary) |
| D2 | Decision latency / governance responsiveness | Speed and quality of decisions once authority exists |
| D3 | Delivery-evidence confidence | Status vs reality; forecast traceability |
| D4 | Scope coherence | Objectives, workstreams, immutable obligations |
| D5 | Vendor / control risk | Binding power; conflict of interest |
| D6 | Political / escalation failure modes | Silence acceptance; escalation routes |
| D7 | Receiving leadership capacity | Funded hours / skill of named receiver |
| D8 | Technical foundation (optional) | Never substitutes for G1–G5 or D1–D7 |
**Suggested 1–5 anchors (educational):** 1 = decision-hostile / unknown-as-green; 3 = mixed / partial evidence; 5 = decision-grade confidence for that dimension. Unknown stays low—do not invent confidence.
How gates + scores combine (no arithmetic override)
Record G1–G5 → display PASS/FAIL.
2. If any critical gate FAILs, list **allowed postures** (often Stop / pause / refer) and **disallowed postures** explicitly.
3. Only then score D1–D8 for education / nuance among still-allowed postures.
4. Composite totals are optional teaching aids; a high total **cannot** unlock a posture blocked by a failed gate.
5. Human qualification still requires authority, sponsor, evidence, access, receiving leadership, capacity, and execution-ownership fit before any Acute engagement.
Hard disclaimer
Gates and scores are decision-support for sponsors. They do **not** qualify a buyer or select an Acute intervention. “Your score = Rescue” language is forbidden.
What evidence changes the call
| Change | Re-run first | Effect |
|—-|—-|—-|
| Authority restored (G1 flips to PASS) | Gates, then dimensions | Continue / Stabilize may become eligible again |
| Evidence unlocked (G2) | Gates, then dimensions | Honest Continue becomes possible; Stabilize access path clears |
| Value≪cost confirmed with no overriding obligation (G3 FAIL) | Gates | Stop / wind-down must be on the table; Continue/Rebuild optimism constrained |
| Showstopper owned by competent authority (G4) | Gates | Refer / Stop vs constrained Continue |
| Receiver named (G5) | Gates | Stabilize / Lead Recovery postures may reopen for ownership discussion |
| Score movement only (D1–D8) | Dimensions among allowed postures | Nuance—not a gate override |
If the receiver is missing, re-run **gates** firstdo not “average up” with high D1–D4 scores.
Independent assessment framing
The incumbent implementer should not be the sole author of continue vs stop. Aligns to **Project Clarity — Diagnose**: diagnose and recommend; client retains execution. If ownership must change later, that is a separate human-qualified Stabilize or Lead Recovery discussion—not a score route.
Worked educational example (labeled EXAMPLE — not a client case)
> **EXAMPLE (fictional composite for teaching). Not a real client. Not a case study. Not a result claim.**
**Situation:** A mid-market ERP post-go-live program reports “green” status. Payroll exceptions are rising. The SI recommends more change orders. Finance cannot produce remaining-value vs remaining-cost that survives scrutiny. The CIO can bind decisions. Evidence access is partial but improving. A delivery director exists on paper but has no funded hours.
**Gates (display first):**
| Gate | Result | Note |
|—-|—-|—-|
| G1 Authority | PASS | CIO + deputy named |
| G2 Evidence access | PASS (fragile) | Exports available; interview access agreed |
| G3 Economic/obligation | **FAIL** | Remaining value clearly below remaining cost; no overriding legal/safety/contractual obligation documented |
| G4 Showstopper | PASS | No unowned legal/safety showstopper |
| G5 Receiver | **FAIL** for Stabilize | “Delivery director” lacks funded capacity |
**Scores (after gates — illustrative only):** Suppose D1–D8 average looks “high” on delivery storytelling (e.g., a teaching total in the upper band). **That total does not matter for unlocking Continue.** G3 FAIL requires Stop / controlled wind-down to be explicitly evaluated; Continue and Rebuild-optimism are constrained. G5 FAIL blocks honest Stabilize until a funded receiver exists.
**Allowed postures in this EXAMPLE:** Stop / controlled wind-down evaluation; pause pending economic truth; specialist referral if regulated rebuild is claimed.
**Disallowed despite “high scores”:** Continue-under-control optimism; Stabilize-as-if-ready.
**Teaching point:** High score + failed gate → constrained posture. Totals never override gates.
FAQ
**Can a score pick Rescue?**
No. Scores do not select Clarity, Rescue, or Recovery.
**Can a high total override a failed gate?**
No.
**Is stop a failed engagement?**
No. A stop recommendation can be a successful independent diagnosis.
**Is this Acute’s service menu?**
No. These are decision postures. Interventions are separate and human-qualified: [Project Clarity — Diagnose](/project-clarity), [Critical Project Rescue — Stabilize](/project-rescue), [Critical Project Recovery — Lead Recovery](/project-recovery).
**Does Acute guarantee recovery if you Continue?**
No. Layer-3 business outcomes are never guaranteed.
Cite / use without buying
Print or copy the gates-first table. Run G1–G5 before any scoring. List unknowns explicitly. Use the EXAMPLE only as a teaching composite—never as Acute performance evidence.
Related pages
[/resources/troubled-technology-project-assessment](https://www.acutepmc.com/resources/troubled-technology-project-assessment)
- [/resources/first-control-window-72-hours](https://www.acutepmc.com/resources/first-control-window-72-hours)
- [/project-clarity](https://www.acutepmc.com/project-clarity) · [/project-rescue](https://www.acutepmc.com/project-rescue) · [/project-recovery](https://www.acutepmc.com/project-recovery)
- [/for-ai-agents](https://www.acutepmc.com/for-ai-agents)
**Contact:** Andrew.Davenport@acutepmc.com
Acute Project Management Consulting LLC (Arizona) ≠ Mumbai lookalike.