Mission
By the end of this chapter, you can design distinct working, technical, and executive reviews that turn current evidence into authorized decisions and owned action.
- Measurable outcome: Design three review types with distinct audiences, evidence, decisions, and outputs, then replace status theater with exception-driven program control.
- Prerequisites: Chapters 13–18; a roadmap and decision log.
- Work product: A review architecture and executive pre-read for Northstar Devices.
- Time: 70–90 minutes.
Before you read: Predict → Commit → Connect
Predict: Why does sending the same 40-slide deck to engineers, operators, and executives usually fail all three audiences?
Commit: Name the one decision your next program review must enable.
Connect: Think of a meeting that felt busy but changed nothing. Which input, decision right, or output was absent?
Match the review to the work
A program review is a control loop: compare evidence with intent, detect variance or learning, make decisions, and update action. Different loops operate at different resolution and cadence. Collapsing them into one meeting produces either technical detail without authority or executive summary without diagnostic depth.
Use at least three review types:
- Working review: owners coordinate near-term execution, close dependencies, update actions, and resolve cross-workstream exceptions.
- Technical review: qualified reviewers examine design, integration, readiness, failure modes, evidence, and technical decisions.
- Executive review: authorized leaders assess outcomes, material risk, scenarios, resources, and cross-organizational trade-offs.
An incident review, control review, or launch-readiness gate may be separate because it has specialized authority and evidence. Do not relabel a status meeting as a technical review if no competent acceptance owner examines the engineering evidence.
Information should flow upward by exception and consequence; decisions and constraints should flow downward without distortion. The executive view is compressed, not sanitized.
Define inputs, decisions, and outputs
Every recurring review needs a contract:
- Purpose: What condition is this forum controlling?
- Audience and rights: Who contributes, who decides, and who only needs the result?
- Entry evidence: What must be current before attendance is useful?
- Agenda by exception: Which variance, uncertainty, or choice deserves synchronous time?
- Exit: What decisions, actions, changed baselines, or escalations must be recorded?
Status that needs no discussion can be read asynchronously. Use synchronous time for interpretation, conflict, and closure. Circulate pre-reads early enough for review, state requested decisions at the beginning, and distinguish facts from forecasts and recommendations.
Working reviews should inspect interfaces, dependency changes, overdue decisions, new evidence, and actions until the next control point. Technical reviews should state acceptance criteria before the review, identify independent or cross-functional reviewers, and preserve findings with disposition. Executive reviews should show the outcome, changed forecast, top material exposures, options, and explicit asks.
Avoid common review failure modes
Round-robin reporting: every lead recites activity, leaving no time for cross-program problems. Replace with pre-read updates and an exception list.
Color without evidence: red, amber, and green vary by speaker. Attach observable thresholds and trend.
Late surprise: a team waits for perfect diagnosis before surfacing risk. Allow preliminary signals labeled with confidence and a next evidence time.
Decision laundering: a recommendation appears approved because no one objected. Record an explicit decision by an authorized owner.
Technical theater: impressive diagrams mask missing test evidence, operability, or consumer acceptance. Review against criteria.
Executive archaeology: leaders search through dozens of slides for the ask. Put material change and decision first.
The TPM improves the signal path but must not edit away justified dissent. If technical reviewers disagree, show the decision owner the disagreement, evidence, and consequence.
Recurring case: Northstar Devices
Northstar's weekly review contains 32 slides, one per function. The supplier lead says samples are amber. Firmware says 85% complete. Regional teams list certification dates. Executives ask whether the September launch is safe, but no one can answer.
The TPM creates three loops. A twice-weekly working review focuses on battery qualification, firmware reproduction, certification dependencies, and owned actions. A technical readiness review examines the production-intent hardware/firmware combination against thermal, upgrade, connectivity, and recovery criteria. A monthly executive review begins with the launch outcome, active scenario, material guardrails, and decisions on region scope, expedited tooling cost, and commitment.
The same source records feed all three. The executive pre-read says: “Reference scenario moved from a two-region September pilot to an October range because production cells breached the approved thermal threshold. Decision requested by Friday: fund alternate-cell qualification and preserve two regions, or hold spend and move to one-region scope.”
Decision rights: Who owns what?
- Workstream owners: provide current evidence, forecasts, and owned recovery actions.
- Technical chair and domain approvers: determine technical readiness or findings within defined authority.
- Sponsor and executives: decide material outcome, funding, scope, or residual business-risk trade-offs.
- Control owners: accept or reject criteria governed by policy or regulation.
- TPM: designs the review system, maintains a trusted source, curates exception signal, prepares decisions, and records outcomes. The TPM does not certify technical readiness without authority or turn an executive preference into an unrecorded command.
I do
I write the purpose and output of each forum. I remove information-only round-robin items from the working agenda and require owners to update the shared record beforehand. I begin with changes since the last review, decisions due before the next, and thresholds breached.
For the executive pre-read, I use five lines: intended outcome; what changed; evidence and confidence; options and consequences; recommendation and ask. I link deeper technical evidence rather than pasting it into the executive layer.
We do
Together, compress this update: “Hardware is amber. We held several supplier calls. Testing continues and the team is working hard.”
We replace it: “Eight of 30 production-intent cells exceeded the approved thermal threshold in the first profile; failure analysis is incomplete. This invalidates the September reference scenario unless an alternate cell begins qualification by July 8. Options: fund expedited qualification, reduce the pilot to the already certified region, or move the range to October–November. Hardware recommends expedited qualification; sponsor decision required Thursday.”
You do
Audit one recurring review. State its purpose, decider, entry evidence, and required output. Remove or move asynchronous items. Then design a linked working, technical, and executive review architecture for your program. Write a one-page executive pre-read centered on one material decision.
Show the model answer
Model answer and 0–4 rubric
Working review purpose: Protect the next two weeks of integration by closing changed dependencies and overdue actions. Inputs: owner updates, dependency graph, decision aging, risk triggers. Output: decisions, named actions, forecast changes, escalations. Technical review purpose: Determine whether the production-intent Northstar configuration is ready for a 1,000-device pilot. Inputs: version manifest, criteria, thermal/firmware/cloud evidence, open defects, recovery test. Output: accept, accept with conditions, or reject; findings and owners. Executive review purpose: Choose the launch scenario and authorize cross-functional trade-offs. Input: outcome, active scenario, material evidence, options. Output: scope/cost/date decision and accepted residual exposure. Executive ask: Decide by Thursday whether to fund alternate-cell qualification or reduce pilot scope; absence of a decision activates the October forecast.
Rubric
- 0 (Missing): Unstructured status meeting with no decision or output.
- 1 (Emerging): Cadences differ, but content and rights remain mixed.
- 2 (Functional): Review purposes, inputs, and outputs are distinct; evidence or escalation discipline is incomplete.
- 3 (Strong): Exception-driven loops share trusted records and route explicit decisions to authorized owners.
- 4 (Decision-ready): Level 3 plus measurable review effectiveness, independent technical challenge, documented dissent, and reliable downward communication of decisions.
Pause & Recall
From memory, explain the purpose of working, technical, and executive reviews. Name three forms of review theater. Connect to Chapter 16: how does a review output enter the decision log?
Production lens
Measure whether reviews reduce decision age, dependency surprises, repeated findings, and action slippage. Periodically sample decisions announced in meetings and ask downstream owners to state them; divergence reveals communication loss. Rotate or add reviewers when independence or expertise is weak. During high-severity incidents, use the incident command cadence rather than forcing routine governance, then reconcile decisions and follow-ups afterward.
Workplace artifact: Review contract and executive pre-read
# Review contract
Purpose / controlled condition:
Cadence and audience:
Decider / technical acceptance owner:
Entry evidence and pre-read deadline:
Exception agenda:
Decisions requested:
Required output:
Where decisions/actions are recorded:
Effectiveness measure:
# Executive pre-read
Outcome:
What changed:
Evidence / confidence:
Impact and active scenario:
Options / recommendation:
Specific ask / owner / deadline:Chapter compression
Working, technical, and executive reviews operate at different resolution and authority. Give each a contract, move static status to pre-reads, use synchronous time for exceptions and decisions, and preserve one traceable evidence-to-decision path.
Retrieval deck
- Q: What is the main output of a working review? A: Closed coordination decisions, owned actions, updated forecasts, and escalations for near-term execution.
- Q: What distinguishes a technical review? A: Qualified review of defined criteria and evidence with an explicit readiness disposition.
- Q: What belongs first in an executive review? A: Outcome, material change, consequence, options, and the decision requested.
- Q: Why is silence not approval? A: Only an explicit choice by an authorized owner creates a traceable decision.
- Q: What should flow upward by exception? A: Material variance, uncertainty, risk, and decisions, rather than every activity detail.
Spaced review
- Now: State the distinct purpose of each review type.
- +1 day: Recreate the three review contracts from memory.
- +3 days: Convert one round-robin agenda into an exception agenda.
- +7 days: Calculate how much of the last review was information-only.
- +14 days: Test whether attendees can state the executive ask before opening the pre-read.