Measure and Keep Improving

Choose evidence that answers the business question and identify what it cannot prove.

Work in progress · 0/2 parts checked.

Understand the situation

Back at Northstar

The revised job runs successfully. That is useful execution evidence, but Northstar agreed on an operating outcome: eligible requests reach accountable sellers.

An observed baseline is measured in your own process. A historical approximation is a partial estimate. An external benchmark is context from somewhere else. A target is what you hope to achieve. Keep those labels visible. When no baseline exists, start by capturing the missing events.

Use a small pilot to find exceptions and user workarounds. After launch, review acceptance, response, and the cost of operating the process. Business changes may make a good design stale. More automation, less automation, or a restored human checkpoint can each be the right next change.

“The job log is green. Did anyone get useful work, and can we tell?”Taylor · Northstar Learning · Teaching case

Supplied sample: Six eligible demo requests and two internal test records. All eight jobs succeeded.
Acceptance evidence: Four eligible requests accepted; one pending; one has no captured acceptance state.
Baseline: Three of six eligible requests accepted in an earlier illustrative window.
Limitation: Small synthetic samples. No experiment establishing causation.

Your GTM Engineering Project Brief

Your recommendation and notes from the example case.

Your notes or a problem you want to try next (optional)

Read the current brief

# GTM Engineering Project Brief

Learning exercise: Northstar Learning is a teaching example. These choices are not a deployed system or a claim of professional certification.

Current status: Draft or changed since validation

## Customer onboarding recommendation

### State the onboarding problem without blaming the team
- Observed problem: Not selected
- Who does this serve?: Not selected

### Choose the smallest recommendation that fixes the observed gap
- First intervention: Not selected
- Operating responsibilities: Not selected
- What still needs investigation?: Not selected

### Choose the evidence that would tell you whether the handoff improved
- Primary measure: Not selected
- Handling missing ownership: Not selected
- Review plan: Not selected

## Learning notes from the inbound case

### Problem brief
Sort what we know from what we are guessing: not currently validated.
Choose what to inspect before buying or building: not currently validated.

### Clients and outcome
Assign each responsibility to a real owner: not currently validated.
Choose the outcome all three teams can live with: not currently validated.

### GTM System Map
Build the five-part system map: not currently validated.
Say exactly what this handoff receipt proves: not currently validated.

### Architecture and decision note
Assign each fact or job to the system that should own it: not currently validated.
Choose a workable intervention and name the tradeoff: not currently validated.
Selected implementation: Not selected.

### Mini workflow contract
Route four requests into the right queue: not currently validated.
Put the recovery steps in the order you would actually run them: not currently validated.

### Measurement and review plan
Build the denominator from the actual records: not currently validated.
Choose the conclusion the evidence supports: not currently validated.

## My notes (not automatically graded)
No personal notes added.

## Source references
- https://gtmjosh.com/reference/gtm-engineer
- https://gtmjosh.com/reference/gtm-engineering-problem-solving
- https://gtmjosh.com/reference/gtm-engineering-vs-revops
- https://gtmjosh.com/reference/gtm-system-anatomy
- https://gtmjosh.com/reference/crm-entities-leads-contacts-accounts-opportunities
- https://gtmjosh.com/reference/gtm-tech-stack-architecture
- https://gtmjosh.com/reference/deterministic-rules-vs-ai-judgment
- https://gtmjosh.com/reference/gtm-engineering-use-cases
- https://gtmjosh.com/reference/gtm-workflow-specification-template
- https://gtmjosh.com/reference/handoff-contracts-between-systems
- https://gtmjosh.com/reference/retry-defer-and-escalation
- https://gtmjosh.com/reference/leading-and-lagging-gtm-indicators
- https://gtmjosh.com/reference/reporting-populations-and-denominators

Look it up in Reference

The explanation and implementation detail, without the exercise.

← Previous chapter Read next chapter →

Replay changes your reading position only. It keeps your drafts and earned history. Editing a checked answer makes that part need validation again.