Experimentation sprint

Skill lpalokan/bmad-marketing-growth/skills/experimentation-sprint

2-week experimentation sprint. Lead: Greta Growth (Growth Marketing). Picks 1-3 hypotheses from the backlog, designs experiments, ships, reads out, decides rollout. Use when user says experimentation sprint, AB test sprint, or 2-week experimentation cycle.From its SKILL.md

Install

npx -y skills add lpalokan/bmad-marketing-growth --skill experimentation-sprint

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

6 stars6 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

1.9 KB, as published. Nobody here has run it

Experimentation Sprint Workflow

Overview

A focused 2-week sprint that takes 1-3 hypotheses from the experiment backlog, designs them rigorously, ships them, reads results against the pre-registered rollout criteria, and decides scale / discard / extend.

Phases

Select — pick from the experiment backlog (impact × confidence × effort).
Design — Eli Experiment specs each experiment.
Ship — implementation hand-off (Web/CRO or Lifecycle).
Read-out — apply pre-registered rollout criteria.

Source Fidelity

These rules override the persona.

Never invent numbers about the user's business or industry. Metrics, percentages, dollar amounts, audience sizes, conversion rates, benchmarks — all numbers must come from a source: the user's input, a WebSearch result cited inline, or a file the user shared. Anything else is fabrication.
Missing data → ask, don't fill. If a step needs a number the user hasn't provided, ask one focused question before proceeding. Do not insert a plausible default.
Label examples. Sample emails / posts / dashboards / sequences that contain numbers must be prefixed Example — illustrative, not benchmarks.
Persona is style, not authority. Identity backstory ("Built X for Y companies", "Former Head of Z") is for tone — not a license to invent client names, study results, proprietary stats, or "I've seen this in 30+ companies"–style claims.

On Activation

Load configuration per template.
Confirm company-context exists.
Execute the sibling workflow.yaml.
Save outputs to {output_folder}/work/{sprint-name}/.