Growth Marketing skills
Growth Marketing Skills-md for Growth Teams, Claude & Codex
Growth marketing skills are SKILL.md files that turn Claude, Codex and other AI agents into careful growth leads. They find the constraint holding growth back, choose a north star and metric tree, build cohort-based growth models, run an experimentation system with properly designed and analysed tests, find new acquisition channels and loops, and build the funnels, cohorts, tracking plans and reports a growth team runs on. All 13 are below: copy one, download it, or take the set.
13 skills, 19,331 words, version 1.0.0, written October 2026.
Growth Audit
Audit a company's growth system end to end: north star and goals, growth model, acquisition, activation, retention, monetisation, referral and loops, data and tracking, experimentation practice and reporting, find the single biggest constraint with evidence, check retention before acquisition and produce a scored, ranked 90-day plan.
North Star Metric and Metric Tree
Choose and validate a north star metric that reflects delivered customer value and predicts revenue, then build the metric tree beneath it: candidate tests, data validation, input metrics with owners and levers, targets, counter-metrics that stop gaming, tracking requirements and a weekly review cadence for SaaS, marketplaces, media, e-commerce, fintech and APIs.
Growth Model and Forecasting
Build cohort-based growth models and forecasts: new users by channel, activation, retention curves, monetisation, expansion, referral and reactivation, calibrated on history, a baseline projection, sensitivity analysis to rank levers, conservative, expected and stretch scenarios tied to initiatives, input targets with owners and monthly actuals versus forecast.
Growth Experiment Backlog and Cadence
Set up and run a growth experimentation system: ideas tied to the metric tree from data, research and support, one backlog, ICE, PIE or RICE prioritisation calibrated by growth model sensitivity, a weekly growth meeting, experiment cards, a tagged learning library and velocity and win-rate tracking.
Experiment Design for Growth
Design trustworthy growth experiments: specific hypotheses, the right design (A/B, A/B/n, holdout, geo or quasi-experiment), randomisation units, primary, secondary and guardrail metrics, sample size, minimum detectable effect and duration, variance reduction, pre-registered stopping rules and segments, and pre-launch QA.
Experiment Readouts and Analysis
Analyse experiment results and write readouts that lead to sound decisions: sample ratio mismatch, exposure, tracking parity and balance checks, effect sizes with intervals, guardrails, secondary metrics, trends and novelty, pre-registered versus exploratory segments, practical significance translated into north star and revenue impact, and a ship, iterate or stop decision.
Acquisition Channel Testing
Find and test the next acquisition channel: the full channel universe across paid, organic and AI search, content and community, partnerships, sales-led, product-led, PR and offline, scoring by reach, fit, cost, speed and scale, cheap time-boxed tests with test cards, tracking for hard-to-measure channels, success judged on activated and retained customers, and scaling winners until marginal CAC nears the limit.
Growth Loops
Identify, design and strengthen growth loops where each user brings more: collaboration and viral loops, referral, user-generated and company-generated content loops, paid and sales loops and product output loops, with loop maps, viral coefficient, content conversion and cycle time, the weakest step, consented mechanics and experiments on each loop step.
Referral Programmes
Design, launch and improve referral programmes for apps, e-commerce, SaaS and B2B: readiness and economics, two-sided incentives, asks timed to moments of satisfaction, placements, one-tap sharing including WhatsApp, the referred friend's experience, fraud controls, referral funnel metrics, referred customer quality and reward tests.
Funnel Analysis
Build and analyse funnels from event data to find where and for whom value leaks: step definitions with windows, step and overall conversion, time between steps, segment breakdowns by device, channel, plan and cohort, opportunity sizing, diagnosis with recordings, surveys and support tickets, evidence-based hypotheses for the backlog and weekly monitoring.
Cohort Retention Analysis
Build and read cohort retention analyses for users, accounts and revenue: value-based activity definitions, N-day, unbounded and bracket retention, logo retention, net revenue retention and repeat purchase rate, cohort tables and curves, flattening, cohort comparisons over time, segments by channel and plan, retention-linked behaviours and actions.
Tracking Plan and Event Taxonomy
Design tracking plans and event taxonomies for product analytics in Amplitude, Mixpanel, PostHog, GA4, Segment, RudderStack or the warehouse: events tied to metrics, object_action naming, event, user and group properties, identity and consent rules, client versus server decisions, audits of messy tracking, QA before release and governance with owners.
Growth Reporting and Attribution
Build growth reporting around the north star and metric tree: a metric dictionary, north star and input views against targets, channel quality by activation and retention, cohort curves, labelled attribution combining last touch, self-reported and test-based evidence, experiment impact, annotations and a weekly growth review that ends with decisions.
No skill matches that search. Clear it or choose All.
On this page
What growth marketing skills do for an AI agent
Ask a general AI model for “growth hacks” and you usually get a list of tactics: launch a referral programme, try TikTok, add exit pop-ups, run more A/B tests. Tactics without a model are guesses. They ignore where growth actually leaks, which lever matters most and whether the team can even measure the result.
These skills work like an experienced growth lead. They start by finding the one constraint holding growth back, usually checking retention before acquisition. They choose a north star that reflects customer value, break it into inputs teams own, and build a cohort-based model that shows which input moves the outcome most. Then they run an experimentation system: a backlog tied to metrics, tests designed with proper samples and metrics, and readouts that check data quality before declaring winners.
Growth marketing, growth hacking and performance marketing
Performance marketing buys measurable traffic; lifecycle marketing keeps customers; growth marketing connects them through models, experiments and analytics across the whole journey. These skills cover the system that decides where effort goes and proves what worked: strategy, experimentation, acquisition and loops, and analytics. They work with Amplitude, Mixpanel, PostHog, GA4, GrowthBook, Optimizely and warehouse-first setups.
Where to start: find the constraint
Start with the growth-audit skill. It reviews goals, the growth model, acquisition, activation, retention, monetisation, loops, data, experimentation and reporting, then names the single biggest constraint with evidence and a 90-day plan.
If the team has no shared goal, set one with north-star-metric-tree. If data disagrees, fix tracking-plan before running experiments. If retention curves never flatten, use cohort-retention-analysis before spending more on acquisition.
Which skill fixes which problem, and the principles
Use this table to go straight to the right skill when you already know the symptom.
| Symptom or task | Likely cause | Skill to use |
|---|---|---|
| Growth stalled and nobody agrees why | No clear constraint or model | growth-audit |
| Teams chase different numbers | No north star or metric tree | north-star-metric-tree |
| Targets feel arbitrary | No cohort-based growth model | growth-model |
| Experiments are ad hoc and slow | No backlog, prioritisation or cadence | growth-experiment-backlog |
| Tests end with no clear answer | Vague hypotheses, underpowered samples | experiment-design |
| Results are disputed | No quality checks or pre-registered analysis | experiment-readout |
| One channel provides most growth | No structured channel testing | acquisition-channel-testing |
| Growth is linear with spend | No loops where users bring users | growth-loops |
| Referral programme underperforms | Wrong timing, rewards or friend experience | referral-programmes |
| Conversion drops and nobody knows where | Funnels not segmented or sized | funnel-analysis |
| Users don’t come back | Retention not analysed by cohort | cohort-retention-analysis |
| Analytics numbers disagree | No tracking plan or event taxonomy | tracking-plan |
| Weekly meetings debate numbers | No north-star-led reporting | growth-reporting |
The principles behind these skills
| Principle | Why it matters |
|---|---|
| Find one constraint at a time | Improving a step that isn’t the bottleneck barely moves growth |
| Retention before acquisition | Users who leave quickly make every acquisition pound leak away |
| Models before targets | Sensitivity analysis shows which input moves the outcome most |
| Power every experiment | Underpowered tests produce noise that looks like answers |
| Check data quality first | Sample ratio mismatch and broken tracking invalidate results |
| Judge channels on retained users | Sign-up volume often ranks channels in the wrong order |
| Label attribution for what it is | Last touch, self-reported and test results each see part of the picture |
Example prompts for each skill
Each skill triggers on everyday wording. These prompts are good starting points:
- Audit: “Sign-ups grow but revenue is flat. Run a growth audit.”
- North star: “Pick our north star metric and build the metric tree.”
- Model: “What would it take to double revenue in 12 months? Build the model.”
- Backlog: “Set up a weekly growth meeting and experiment backlog.”
- Design: “How many users does our onboarding test need?”
- Readout: “Analyse this A/B test and tell me whether to ship.”
- Channels: “Google brings 70% of customers. Find the next channel.”
- Loops: “Map our growth loops and find the weakest step.”
- Referral: “Our referral programme gets almost no shares. Fix it.”
- Funnels: “Find where and why sign-ups fail to activate.”
- Retention: “Build cohort curves and tell me if we have product-market fit.”
- Tracking: “Clean up our 340 analytics events.”
- Reporting: “Build a weekly growth report that ends in decisions.”
What the agent needs from you
Growth work depends on data. The most useful inputs are:
- Event data from analytics or the warehouse, with user or account IDs and timestamps.
- Revenue and pricing by month and cohort.
- Acquisition data by channel with downstream activation and retention.
- Cohort retention curves or the data to build them.
- Experiment history: what was tested, results and decisions.
- Customer research: interviews, surveys, support tickets and recordings.
If something is missing, the skill works with what you have and says which conclusions would change with more data.
What an agent cannot do, and myths to ignore
An agent can analyse, model, design and write, but it can’t see your analytics unless you export or connect it, and it can’t run experiments in your product on its own. Strategic choices, such as which segment to focus on or which bet to fund, stay with you; the skills give you evidence and trade-offs.
| Myth | What the evidence says |
|---|---|
| Growth hacks beat fundamentals | Lasting growth comes from retention, models and steady experimentation |
| More experiments always help | Underpowered tests waste time and mislead |
| Sign-ups are growth | Users who don’t retain add cost, not growth |
| Last-touch attribution shows what works | It over-credits bottom-funnel channels; combine with self-reported and test evidence |
| A significant result is always worth shipping | Check data quality, guardrails and practical significance first |
Frequently asked questions
What is growth marketing?
A discipline that uses models, experiments and analytics across the whole customer journey (acquisition, activation, retention, monetisation and referral) to find and prove what drives sustainable growth.
How is growth marketing different from performance marketing?
Performance marketing buys measurable traffic. Growth marketing decides where effort should go across the journey and proves what worked, including product, lifecycle and loop changes.
What is a north star metric?
The one metric that best reflects the value customers get and predicts revenue, broken into input metrics teams can move.
How many users does an A/B test need?
It depends on the baseline rate and the smallest effect worth detecting. A 5% baseline and a 10% relative change often need around 30,000 users per variant.
What is a growth loop?
A system where each cycle’s output becomes the next cycle’s input, such as users inviting colleagues who invite more, so growth compounds.
How do I know if I have product-market fit?
One strong signal is cohort retention curves that flatten at a healthy level for a segment: the people who stay, keep staying.
Which analytics tools do these skills support?
Any: Amplitude, Mixpanel, PostHog, GA4, Segment, RudderStack and warehouse setups, plus experimentation tools such as GrowthBook and Optimizely.
How do these skills relate to the other marketing skills?
Growth skills set the goals, models and experiments; Performance Marketing, CRM and Lifecycle, SEO and AEO skills execute in each channel.
Sources and more marketing skills
- Amplitude North Star Playbook
- Kohavi, Tang and Xu, Trustworthy Online Controlled Experiments (Cambridge University Press, 2020)
- Gabriel Weinberg and Justin Mares, Traction (2015)
- Evan Miller sample size calculator
- GrowthBook documentation
Checked on 10 October 2026. Pair this set with the Performance Marketing skills for paid acquisition and the CRM and Lifecycle Marketing skills for retention. See every subcategory on the Marketing Skills-md page.
