Milestone map
Milestone map
3 milestones
Define the Measurement Framework for a Real Campaign
1–2 weeks
Select a real marketing campaign you have access to performance data for — either your own organisation's campaign or a publicly available campaign with documented results. Define the measurement framework: campaign objective (awareness, consideration, conversion), key performance indicators, attribution model used, and the counterfactual (what baseline are you comparing against).
Proof required
Submit: (a) campaign brief summary (one page) describing the campaign objective, target audience, channels used, and budget (approximate if confidential), (b) measurement framework document specifying the KPIs, attribution model, and baseline for comparison, and (c) confirmation that you have access to the actual performance data for this campaign (access to real data is required — this outcome cannot be satisfied with hypothetical data).
What gets checked
- Campaign brief is for a real campaign with real performance data — not a hypothetical
- Measurement framework specifies the attribution model explicitly — last click, multi-touch, or data-driven attribution produce very different results
- Baseline for comparison is stated — without a counterfactual, campaign performance cannot be assessed
Common mistakes
- Using a case study from a textbook rather than a real campaign with real data
- Attribution model not specified — 'we tracked conversions' without specifying which touchpoints received credit is not a measurement framework
Powstik Guide
A measurement framework for one real campaign.
Steps
- Summarise the campaign on one page.
- Set KPIs tied to its objective.
- Name the data source for each KPI.
- Set a target for each KPI.
Template
Campaign: Objective: Audience: Channels: | KPI | Target | Data source | |-----|--------|-------------|
What gets sent back
- A hypothetical campaign.
- KPIs not linked to the objective.
- No data sources.
Resources
Foundationstart here
Depthgo deeper
What a verifier looks for
- Verify that the submitter has access to real campaign data — ask them to share a sample data table or screenshot with any sensitive data redacted.
- Check the attribution model specification: can the submitter explain what touchpoints receive credit and why that model was chosen over alternatives?
You'll sign in first, then come straight back here.
Analyse Campaign Performance and Produce Findings
2–3 weeks
Conduct the campaign performance analysis: compute all specified KPIs, compare against baseline, segment performance by channel and audience where data permits, and produce findings on what drove performance. The analysis must go beyond headline metrics to explain why performance was what it was.
Proof required
Submit a campaign performance analysis (minimum 1,500 words) covering: (a) KPI performance table showing actuals vs. targets for all specified metrics, (b) channel and audience segment performance comparison (minimum three segments), (c) key drivers analysis explaining the top two or three factors that drove performance above or below target, and (d) data limitations section acknowledging measurement gaps and their effect on conclusions.
What gets checked
- KPI table shows actuals vs targets — not just actuals
- Driver analysis explains causation, not just correlation — 'social media drove 40% of conversions' needs an explanation of why
- Data limitations are specific — 'we cannot distinguish brand vs performance attribution in the display channel because view-through windows are uncertain'
Common mistakes
- Analysis reports headline metrics without segmentation — a campaign that performed at target on average may have significantly over- and under-performed by segment
- Driver analysis assigns causation without evidence — correlation between channel exposure and conversion does not confirm the channel caused the conversion
Powstik Guide
A 1,500-word campaign analysis with a KPI table.
Steps
- Fill actual vs target for each KPI.
- Explain the biggest gaps.
- Compare channels on cost per result.
- Recommend what to change next time.
Template
| KPI | Target | Actual | Gap | Why | |-----|--------|--------|-----|-----| Channel comparison: Recommendations:
What gets sent back
- Under 1500 words.
- Claims without sources.
- No clear recommendation.
Resources
Foundationstart here
Depthgo deeper
What a verifier looks for
- Challenge the driver analysis: 'You attributed the above-target conversion rate to the email channel — but email and social ran simultaneously. How do you know email drove the lift rather than social?'
- Check the data limitations section: is it specific to this campaign's data, or generic?
You'll sign in first, then come straight back here.
Present Analysis and Recommendations to a Marketing Analytics Professional
1 week
Present the campaign performance analysis and recommendations for future campaigns to a marketing analytics professional (media analyst, marketing strategist, or digital marketing manager) who will challenge your attribution assumptions or driver conclusions.
Proof required
Submit: (a) a recommendations document (minimum 500 words) for the next campaign based on your analysis, (b) a Q&A log from the review session (minimum 300 words with at least two challenges and your responses), and (c) attendance record with reviewer name, credentials, and date.
What gets checked
- Recommendations are specific and grounded in the analysis — not generic marketing best practices
- Q&A log shows the reviewer challenged attribution or causal reasoning
- Reviewer has marketing analytics expertise
Common mistakes
- Recommendations are general ('invest more in social media') without specific targets or rationale from the data
- Review session with someone who approves everything without challenge
Powstik Guide
Next-campaign recommendations, tested by a marketing analyst.
Steps
- Turn each finding into one recommendation with a target number.
- State the attribution model you used and its blind spot.
- Book a media analyst, strategist or digital marketing manager.
- Ask them to attack your attribution and causal claims.
- Log the session and adjust the recommendations.
Template
Recommendation table | Finding (with number) | Recommendation | Target | How we will measure | |-----------------------|----------------|--------|---------------------| Review session log Date: Duration: Reviewer name: Reviewer role and years of experience: Challenge 1 (attribution or causality): What they asked: My answer: Did it hold? What would I change? Challenge 2: What they asked: My answer: Did it hold? What would I change? Change I made after the session: Reviewer's overall view (their words):
What gets sent back
- Generic advice like 'invest more in social'.
- No challenge to the attribution in the log.
- A reviewer without marketing analytics experience.
Resources
Foundationstart here
Depthgo deeper
Masteryfor the dedicated
What a verifier looks for
- Challenge the recommendations: 'You're recommending increasing the email budget — but your own analysis showed email had the highest unsubscribe rate during this campaign. How do you reconcile that?'
- Provide a written confirmation (minimum 150 words) of your marketing analytics credentials and the challenges you raised.
You'll sign in first, then come straight back here.