Milestone map
Milestone map
3 milestones
Run a Campaign with a Measurable Outcome
4–6 weeks (planning + execution + reporting)
Design and execute a real marketing campaign — paid, organic, or email — with a specific goal and a specific audience. The campaign must have a defined start date, end date, and a numeric goal established before launch (e.g. acquire 50 email subscribers, generate 20 trial signups, achieve 3% email click-through rate, or reach 1,000 organic impressions for a specific post). The campaign must run for at least 2 weeks with at least 3 distinct pieces of content or ad creatives.
Proof required
Submit the campaign brief (goal, audience, channels, timeline, budget if any), screenshots of the 3+ creatives or content pieces before launch, and the performance report: actual vs target metric, the highest-performing creative and why you think it performed, and one specific hypothesis for the next campaign.
What gets checked
- Numeric goal was established before launch — a goal defined after the results are known is a description, not a target
- Performance report names the specific metric and the actual number (not 'the campaign performed well') — '42 signups vs 50 target: 84% of goal' is the correct format
- The hypothesis for the next campaign is falsifiable ('I believe using [specific element] will increase click rate by 10% because [reason]') not aspirational ('do better next time')
Resources
Foundationstart here
Depthgo deeper
What a verifier looks for
- Ask the submitter to state the campaign goal and actual result in one sentence — if they cannot do this without checking their notes, they were not tracking the campaign in real time
- Check that the goal was established before launch — ask for the campaign brief with a date predating the first creative
- Ask the submitter what they would test in the next campaign and why — the hypothesis must be specific and falsifiable
- Confirm the campaign reached real people — ask for an impression count or send count from the analytics; a campaign of fewer than 100 people has no reliable performance signal
- A campaign that did not reach its target audience — a campaign is only as good as its distribution; if fewer than 100 people saw it, the performance data is not meaningful
- 3 creatives that are identical except for minor copy variations — the minimum viable test requires at least 2 meaningfully different approaches to learn anything
- A performance report that attributes the result to external factors without identifying anything actionable — 'the algorithm changed' is not a hypothesis for the next campaign
Optimise a Channel Based on Performance Data
10–12 weeks (2 months baseline + 8 weeks optimisation)
Take one marketing channel — email, paid search, organic social, or SEO — and improve its performance by making evidence-based changes over 8 weeks. Start with a baseline measurement (what are the key metrics now?), form one specific hypothesis about what change will improve performance, implement the change, and measure the result at the end of 8 weeks. The channel must already exist with at least 2 months of historical data.
Proof required
Submit the baseline data (key metric and value from the 2-month historical period), your hypothesis (if I change X to Y, metric Z will improve because [reason]), the specific change you made (with before/after screenshots if applicable), and the 8-week result report (metric value after the change, confirmation or rejection of the hypothesis, and the next hypothesis).
What gets checked
- Hypothesis is specific and falsifiable — 'if I change the email subject line from first-person to question format, open rate will increase by at least 5 percentage points' is testable; 'better subject lines will improve opens' is not
- 8-week result report explicitly states whether the hypothesis was confirmed or rejected — not whether the metric went up or down, but whether the specific hypothesis was correct
- The next hypothesis is informed by the result — if the hypothesis was rejected, the next hypothesis addresses why it might have been wrong
Resources
Foundationstart here
Depthgo deeper
What a verifier looks for
- Ask the submitter to state the hypothesis verbatim from their document before showing any results — this confirms it was written before the result was known
- Check that only one variable changed during the 8-week period — ask the submitter to confirm no other significant changes were made to the channel during the test
- Ask the submitter whether the hypothesis was confirmed or rejected and how they reached that conclusion — the reasoning is more important than the binary outcome
- Ask for the next hypothesis and whether it was informed by the result or would have been the same regardless — genuine optimisation learning produces hypotheses that could not have been written before the test
- Making multiple changes at the same time and attributing the result to one of them — changing subject line, send time, and content in the same week makes it impossible to identify which change drove the result
- A hypothesis that is a directional wish ('I expect it to improve') rather than a specific falsifiable prediction with a threshold
- Measuring a channel with too few data points to detect a real change — a newsletter with 50 subscribers has too small a sample to detect a 5% open rate change reliably
Present Your Marketing Portfolio to a Senior Marketing Professional
1–2 weeks (60–90 min session)
Present your M1 campaign and M2 optimisation to a qualified reviewer: a senior marketing manager, head of growth, or VP marketing with 5+ years of industry marketing experience. The reviewer challenges the goal choice (why this metric and not [alternative metric]?), the attribution approach (how do you know the campaign caused the result?), and the next hypothesis (what assumptions are you making that might be wrong?). The reviewer also presents a real marketing brief — a business goal they name — and asks you to design a channel + campaign approach live.
Proof required
Submit your M1 and M2 artefacts plus the reviewer's written assessment: the goal challenge and response, the attribution challenge and response, the hypothesis assumption challenge and response, the marketing brief and your live channel + campaign design, and their sign-off confirming their role and a live session.
What gets checked
- Attribution challenge response acknowledges the limits of attribution (correlation vs causation, multi-touch) not just a defence of the measurement approach
- Live marketing brief response names a specific channel, a specific format, and a specific numeric goal — not a general strategy statement
- Reviewer sign-off confirms 5+ years of marketing in industry (not academic or agency-only)
Resources
Foundationstart here
What a verifier looks for
- The attribution challenge is the most important question — ask the submitter what evidence would convince them the campaign did NOT cause the result; this tests whether they understand correlation vs causation
- The marketing brief should be a real business goal (increase enterprise trial signups, launch in a new geography, defend a shrinking category) — it should require genuine strategic thinking about channel selection
- Reviewer minimum qualification: 5+ years of marketing with direct accountability for performance metrics; brand or communications backgrounds without performance marketing experience are less appropriate
- Ask the submitter what they know about marketing now that they did not know before M1 — this tests whether the work produced genuine conceptual learning, not just execution experience
- Attribution challenge response that defends the measurement approach without acknowledging its limits — no marketing attribution is perfect and a reviewer who tests this knows it
- Marketing brief response that describes what to say without choosing a specific channel — channel selection is the primary marketing strategy decision
- A reviewer who is primarily a brand or communications professional without digital campaign performance experience — the challenge questions require someone who has owned performance metrics