In Part 1, we covered the OP1/OP2 framework. In Part 2, we wrote SMART goals and team tenets. In Part 3, we traced every goal from the keyboard to the P&L. In Part 4, we decomposed goals into resources and cost.
All of that answers the question: what should we work on, and what does it cost? But there's a bigger question most teams never confront: what happens after goals are set?
The honest answer, in most organisations, is: not much. Goals get set in January and forgotten by March. They surface again at year-end when someone needs to write a performance review. Setting goals is 20% of the work — tracking, reviewing, and course-correcting is the other 80%.
This post covers the operational rhythm that keeps goals alive: the review cadences (Weekly Flash, MBR, QBR, Annual Review), what to do when goals go off-track, the TPM/Program Manager role that orchestrates all of it, how AI tools can streamline the process, and a fully interactive dashboard to monitor everything.
The Review Cadence Pyramid
Goal tracking operates on four interlocking cadences. Think of them as a pyramid: the base runs frequently and catches issues early; the apex runs rarely and makes strategic decisions. Each layer feeds the one above it.
| Cadence | Frequency | Audience | Purpose | Duration |
|---|---|---|---|---|
| Weekly Flash | Weekly | Team + skip-level | Surface blockers fast, keep momentum visible | 15 min async read |
| MBR | Monthly | Department + cross-functional partners | Deep-dive into metrics, resource reallocation, dependency resolution | 60 min |
| QBR | Quarterly | VP/Director + Finance + HR | Goal health vs. plan, budget reforecast, hiring pipeline check | 90 min |
| Annual Review | Annually | C-suite + Board | Year-end actuals vs. OP1 commitments, next year's OP1 inputs | Half-day |
For each cadence, someone has to own the rhythm: setting the agenda, collecting inputs, running the meeting (or publishing the async doc), and tracking follow-ups. That someone is the TPM or Program Manager — more on that role in a later section.
Common mistakes across all cadences
Too long. Weekly flashes that require a 30-minute read defeat the purpose. MBRs that run 2 hours lose the room. Brevity forces clarity.
No RAG status. Without a red/amber/green signal, everything looks fine until it isn't. RAG is a forcing function for honesty.
Burying the bad news. If the red items are on slide 47, you have a culture problem, not a formatting problem. Lead with what's broken.
No action items. A review without actions is a book club. Every review should produce a short list of decisions made and actions assigned.
The Weekly Flash (Deep Dive)
The Weekly Flash is the heartbeat of goal tracking. It's not a meeting — it's a compact, async document that takes 15 minutes to read. Every goal gets a single row with six columns:
| Goal | RAG | This Week | Next Week | Blockers | ETA |
|---|---|---|---|---|---|
| Migrate to K8s | Green | Deployed staging cluster; ran load tests at 2x prod traffic | Begin canary rollout for auth service | — | Aug 15 |
| Launch ML Recommender | Amber | Model accuracy at 78% (target: 85%); investigating feature gaps | Add purchase-history features; retrain | Data pipeline backfill delayed by 3 days | Sep 30 (was Sep 15) |
| SOC 2 Certification | Red | Auditor flagged 3 critical findings in access controls | Path-to-green session scheduled Thursday | Need VP decision on SSO mandate by Friday | At risk — TBD |
RAG Status Definitions
The critical distinction between Amber and Red: Amber means the team has a plan to recover. Red means they need help from above. If a goal is amber for more than two consecutive weeks without improving, it's probably red and someone is being optimistic.
The TPM's Role in the Weekly Flash
- Collect updates async — chase goal owners who haven't submitted by the deadline (typically Wednesday afternoon for a Thursday publish)
- Flag missing updates — a missing update is a red flag in itself; escalate silence
- Collate into one document — the flash should be a single email, doc, or Slack post — not scattered across tools
- Highlight reds at the top — don't bury them. The VP should see reds before they see greens
- Escalate blockers — if a blocker can't be resolved at the team level, the TPM raises it to the right decision-maker
AI opportunity
Auto-draft the weekly flash from Jira/Linear ticket status, Slack threads, and PR activity. The TPM reviews and edits, but the first draft is assembled automatically — saving hours of chasing and copy-pasting.
MBR & QBR Best Practices
Monthly Business Review (MBR)
The MBR is metrics-led, not status-led. Show the numbers first (actual vs. target), then explain the delta. Status updates without metrics are opinions; metrics without context are noise. You need both, in that order.
What goes into an MBR:
- Metrics dashboard — actual vs. target for each goal, with trend arrows
- Delta analysis — for every metric that's off-plan, a brief root-cause explanation
- Resource utilisation — how much of allocated HC/budget has been consumed vs. plan
- Cost-to-date — actual spend vs. budget, burn rate projection
- Dependency map — cross-team dependencies and their status
- Actions from last MBR — what was agreed, what got done, what didn't
Quarterly Business Review (QBR)
The QBR adds a strategic lens. Where the MBR asks "are we executing the plan?", the QBR asks "is the plan still the right plan?" Has the market changed? Has the business context shifted? Should we reforecast the remaining quarters?
What goes into a QBR:
- Goal health summary — RAG status for all goals, with quarter-over-quarter trend
- Budget reforecast — updated projections for the remaining quarters
- Hiring pipeline check — approved HC vs. filled vs. in-pipeline
- Goal portfolio decisions — what to cut, add, or accelerate
- Strategic context — market shifts, competitor moves, regulatory changes that affect goals
- Escalations for executive decision — trade-offs that require VP/Director-level calls
AI opportunity
Auto-generate MBR decks from weekly flashes + dashboards. The AI aggregates four weeks of flash data, pulls the latest metrics, and produces a first-draft deck. It can also surface anomalies — goals that have been amber for 3+ weeks, or metrics moving in the wrong direction that nobody flagged.
When Goals Go Red: Path to Green
A red goal isn't a failure — it's a signal. The failure is ignoring it. "Path to green" is a concrete, time-boxed recovery plan with owners and milestones. It's the difference between "we're behind" and "here's exactly how we get back on track."
Path to Green Template
- Rotate all shared credentials and implement per-user service accounts — Owner: SRE Lead, by Jul 18
- Enable MFA on VPN for all staff — Owner: IT Manager, by Jul 22
- Deploy automated quarterly access review via Okta — Owner: Security Eng, by Aug 5
- Schedule re-audit with auditor — Owner: TPM, by Aug 12
When to Downgrade vs. Cut a Goal
Not every red goal should be recovered. Sometimes the right decision is to cut it — if the business context has changed, if the cost of recovery exceeds the value, or if resources are better deployed elsewhere. The QBR is the right forum for this decision, not the weekly flash.
Downgrade when the goal is still valuable but the original target was too aggressive. Redefine "green" with a revised, achievable target. Cut when the goal no longer serves the strategy. Cutting isn't failure — it's resource discipline.
The Program Manager's Role in Goal Tracking
The TPM (Technical Program Manager) is the operating system for goal execution. They don't own the goals — the engineering and product leaders do. But they own the process that ensures goals get tracked, reviewed, and course-corrected.
Key responsibilities:
- Owns the cadence calendar — sets up recurring reviews, sends agendas in advance, drives follow-ups after
- Maintains the single source of truth — the goal tracker or dashboard that everyone references
- Identifies cross-team dependencies and conflicts — spots when Team A's goal depends on Team B's deliverable, and nobody has talked about it
- Escalates — not solves — blockers — routes the right problem to the right decision-maker, with the right context
- Translates between engineering speak and business speak — turns "we're refactoring the data layer" into "customer-facing latency will improve by 40% in Q3"
- Drives the path-to-green process — facilitates root-cause analysis, writes the recovery plan, tracks the actions, reports back at the next cadence
TPM vs. Project Manager
This is not a project manager with a Gantt chart. A TPM understands the technical domain deeply enough to challenge estimates, spot risks that engineers downplay, and hold leaders accountable for commitments. They're the person who asks "have you actually load-tested that?" and knows when the answer is hand-waving.
Using AI to Streamline Goal Tracking
AI won't replace the TPM or the review cadences, but it can eliminate the drudgery that makes them hard to sustain. Here are five specific, practical use cases:
Auto-Status Collection
Pull ticket status from Jira/Linear, summarise Slack threads, and draft the weekly flash. The TPM reviews and edits rather than assembling from scratch.
Gap Detection
Compare planned vs. actual velocity each week. Flag goals drifting amber before the human notices — catching drift at week 3 instead of week 8.
MBR/QBR Deck Generation
Auto-populate slides from dashboard data and weekly flashes. Surface anomalies: goals amber for 3+ weeks, metrics moving the wrong way, budget overruns.
Path-to-Green Drafting
Given the gap between current state and target, AI drafts a recovery plan outline with suggested actions. The TPM refines with domain knowledge and context.
A fifth use case deserves its own mention: natural-language querying. Instead of filtering spreadsheets, ask "Which goals are red this quarter?" or "What's the total HC allocated to revenue goals?" and get an instant answer. The dashboard below gives a taste of what this looks like.
Planning vs. Tracking
The Goal Simulator from Part 4 helps at planning time — decomposing goals into resources and cost. The tools described here help at tracking time — monitoring execution, catching drift, and generating reports. Both are part of a complete AI-assisted goal management workflow.
The Interactive Goal Dashboard
Below is a fully interactive mock dashboard using sample data. It demonstrates what a real goal tracking dashboard looks like — with filtering, grouping, sorting, and aggregate stats. Try it: search for a goal, filter by RAG status, group by org, or sort by due date.