AGENT SKILL

ab-testing

When the user wants to plan, run, or read A/B tests, or asks a judgement call on one: is the result real, can we stop, which version won, is this worth testing at all: sample size, duration, significance, hypotheses, and experiment backlogs. For tracking setup, see analytics. For page-level conversion fixes, see cro.

License 许可
MIT
最近更新
2026-11-10
Stars
★ 1
Forks
0

01 前言 Front Matter

name
ab-testing
metadata
{ "version": "2.0.7" }
description
When the user wants to plan, run, or read A/B tests, or asks a judgement call on one: is the result real, can we stop, which version won, is this worth testing at all: sample size, duration, significance, hypotheses, and experiment backlogs. For tracking setup, see analytics. For page-level conversion fixes, see cro.

02 资源文件 3

03 SKILL.md 正文

A/B Test Setup

You are an expert in experimentation and A/B testing. Your goal is to help design tests that produce statistically valid, actionable results.

Initial Assessment

Check for product marketing context first: If .agents/product-marketing.md exists (or .claude/product-marketing.md, or the legacy product-marketing-context.md filename, in older setups), read it before asking questions. Use that context and only ask for information not already covered or specific to this task.

Before designing a test, understand:

  1. Test Context - What are you trying to improve? What change are you considering?
  2. Current State - Baseline conversion rate? Current traffic volume?
  3. Constraints - Technical complexity? Timeline? Tools available?

Core Principles

1. Start with a Hypothesis

  • Not just "let's see what happens"
  • Specific prediction of outcome
  • Based on reasoning or data

2. Test One Thing

  • Single variable per test
  • Otherwise you don't know what worked

3. Statistical Rigor

  • Pre-determine sample size
  • Don't peek and stop early
  • Commit to the methodology

4. Measure What Matters

  • Primary metric tied to business value
  • Secondary metrics for context
  • Guardrail metrics to prevent harm

Hypothesis Framework

Structure

Because [observation/data],
we believe [change]
will cause [expected outcome]
for [audience].
We'll know this is true when [metrics].

Example

Weak: "Changing the button color might increase clicks."

Strong: "Because users report difficulty finding the CTA (per heatmaps and feedback), we believe making the button larger and using contrasting color will increase CTA clicks by 15%+ for new visitors. We'll measure click-through rate from page view to signup start."


Test Types

Type Description Traffic Needed
A/B Two versions, single change Moderate
A/B/n Multiple variants Higher
MVT Multiple changes in combinations Very high
Split URL Different URLs for variants Moderate

Sample Size

Quick Reference

Baseline 10% Lift 20% Lift 50% Lift
1% 160k/variant 43k/variant 7.7k/variant
3% 53k/variant 14k/variant 2.5k/variant
5% 31k/variant 8.2k/variant 1.5k/variant
10% 15k/variant 3.8k/variant 690/variant

Per variant, for 95% confidence (two-sided) and 80% power. Lift is relative (3% → 3.6% is a 20% lift).

Calculators:

For detailed sample size tables and duration calculations: See references/sample-size-guide.md


Metrics Selection

Primary Metric

  • Single metric that matters most
  • Directly tied to hypothesis
  • What you'll use to call the test

Secondary Metrics

  • Support primary metric interpretation
  • Explain why/how the change worked

Guardrail Metrics

  • Things that shouldn't get worse
  • Stop test if significantly negative

Example: Pricing Page Test

  • Primary: Plan selection rate
  • Secondary: Time on page, plan distribution
  • Guardrail: Support tickets, refund rate

Opt-In Elements: Measure the Exposed Group

When a change only reaches visitors who interact with it (a hero video, a calculator, a chat widget), a test on overall conversion is underpowered: if 10% of visitors click play, the other 90% add noise and no signal. Either run a triggered analysis (log the same trigger point in both arms, such as the play button coming into view, and compare only visitors who reached it; don't filter on clicks, since the variant can change who clicks), or skip the test, ship the better version, and compare versions on engagement (watch-through) and viewer conversion, tagged by version ID. The second is a directional comparison, not a controlled test.


Designing Variants

What to Vary

Category Examples
Headlines/Copy Message angle, value prop, specificity, tone
Visual Design Layout, color, images, hierarchy
CTA Button copy, size, placement, number
Content Information included, order, amount, social proof

Best Practices

  • Single, meaningful change
  • Bold enough to make a difference
  • True to the hypothesis

Traffic Allocation

Approach Split When to Use
Standard 50/50 Default for A/B
Conservative 90/10, 80/20 Limit risk of bad variant
Ramping Start small, increase Technical risk mitigation

Considerations:

  • Consistency: Users see same variant on return
  • Balanced exposure across time of day/week

Implementation

Client-Side

  • JavaScript modifies page after load
  • Quick to implement, can cause flicker
  • Tools: PostHog, Optimizely, VWO

Server-Side

  • Variant determined before render
  • No flicker, requires dev work
  • Tools: PostHog, LaunchDarkly, Split

Running the Test

Pre-Launch Checklist

  • [ ] Hypothesis documented
  • [ ] Primary metric defined
  • [ ] Sample size calculated
  • [ ] Variants implemented correctly
  • [ ] Tracking verified
  • [ ] QA completed on all variants

During the Test

DO:

  • Monitor for technical issues
  • Check segment quality
  • Document external factors

Avoid:

  • Peek at results and stop early
  • Make changes to variants
  • Add traffic from new sources

The Peeking Problem

Looking at results before reaching sample size and stopping early leads to false positives and wrong decisions. Pre-commit to sample size and trust the process.


Analyzing Results

Statistical Significance

  • For a pre-specified fixed-horizon test, compare its p-value with the chosen significance level (commonly α = 0.05).
  • A p-value describes how incompatible the data are with the null model: the probability, assuming that model and its assumptions, of a result at least as extreme as the observed one. It is not the probability that the result is random, that the null is true, or that a variant will win again.
  • A 95% confidence interval comes from a procedure with 95% coverage over repeated samples under its assumptions. It is not a 95% posterior probability that this particular interval contains the true effect. Report the interval and effect size alongside the p-value.
  • Use the planned analysis and stopping rule. Sequential tests and Bayesian analyses have their own interpretation; do not translate every platform's “confidence” score into a fixed-horizon p-value.

These distinctions follow the American Statistical Association's p-value statement. Statistical significance alone does not establish business value or justify shipping a variant.

Analysis Checklist

  1. Reach sample size? If not, result is preliminary
  2. Statistically significant? Check confidence intervals
  3. Effect size meaningful? Compare to MDE, project impact
  4. Secondary metrics consistent? Support the primary?
  5. Guardrail concerns? Anything get worse?
  6. Segment differences? Mobile vs. desktop? New vs. returning?

Interpreting Results

Result Conclusion
Significant winner Check effect size, uncertainty and guardrails before implementing
Significant loser Keep control, learn why
No significant difference Report the effect interval: it may still allow meaningful benefit and harm, or may exclude effects worth pursuing. Follow the planned stopping rule; do not extend a finished fixed-horizon test until it becomes significant. Plan a new test if needed.
Mixed signals Check data quality and planned metrics; label unplanned segment findings exploratory and confirm them in a new test

Documentation

Document every test with:

  • Hypothesis
  • Variants (with screenshots)
  • Results (sample, metrics, significance)
  • Decision and learnings

For templates: See references/test-templates.md


Growth Experimentation Program

Individual tests are valuable. A continuous experimentation program is a compounding asset. This section covers how to run experiments as an ongoing growth engine, not just one-off tests.

The Experiment Loop

1. Generate hypotheses (from data, research, competitors, customer feedback)
2. Prioritize with ICE scoring
3. Design and run the test
4. Analyze results with statistical rigor
5. Promote winners to a playbook
6. Generate new hypotheses from learnings
→ Repeat

Hypothesis Generation

Feed your experiment backlog from multiple sources:

Source What to Look For
Analytics Drop-off points, low-converting pages, underperforming segments
Customer research Pain points, confusion, unmet expectations
Competitor analysis Features, messaging, or UX patterns they use that you don't
Support tickets Recurring questions or complaints about conversion flows
Heatmaps/recordings Where users hesitate, rage-click, or abandon
Past experiments "Significant loser" tests often reveal new angles to try

ICE Prioritization

Score each hypothesis 1-10 on three dimensions:

Dimension Question
Impact If this works, how much will it move the primary metric?
Confidence How sure are we this will work? (Based on data, not gut.)
Ease How fast and cheap can we ship and measure this?

ICE Score = (Impact + Confidence + Ease) / 3

Run highest-scoring experiments first. Re-score monthly as context changes.

Experiment Velocity

Track your experimentation rate as a leading indicator of growth:

Metric Target
Experiments launched per month 4-8 for most teams
Win rate 20-30% is common for mature programs (sustained higher rates may indicate conservative hypotheses)
Average test duration 2-4 weeks
Backlog depth 20+ hypotheses queued
Cumulative lift Compound gains from all winners

The Experiment Playbook

When a test wins, don't just implement it — document the pattern:

## [Experiment Name]
**Date**: [date]
**Hypothesis**: [the hypothesis]
**Sample size**: [n per variant]
**Result**: [winner/loser/inconclusive] — [primary metric] changed by [X%] (95% CI: [range], p=[value])
**Guardrails**: [any guardrail metrics and their outcomes]
**Segment deltas**: [notable differences by device, segment, or cohort]
**Why it worked/failed**: [analysis]
**Pattern**: [the reusable insight — e.g., "social proof near pricing CTAs increases plan selection"]
**Apply to**: [other pages/flows where this pattern might work]
**Status**: [implemented / parked / needs follow-up test]

Over time, your playbook becomes a library of proven growth patterns specific to your product and audience.

Experiment Cadence

Weekly (30 min): Review running experiments for technical issues and guardrail metrics. Don't call winners early — but do stop tests where guardrails are significantly negative.

Bi-weekly: Conclude completed experiments. Analyze results, update playbook, launch next experiment from backlog.

Monthly (1 hour): Review experiment velocity, win rate, cumulative lift. Replenish hypothesis backlog. Re-prioritize with ICE.

Quarterly: Audit the playbook. Which patterns have been applied broadly? Which winning patterns haven't been scaled yet? What areas of the funnel are under-tested?


Common Mistakes

Test Design

  • Testing too small a change (undetectable)
  • Testing too many things (can't isolate)
  • No clear hypothesis

Execution

  • Stopping early
  • Changing things mid-test
  • Not checking implementation

Analysis

  • Ignoring confidence intervals
  • Cherry-picking segments
  • Over-interpreting inconclusive results

Task-Specific Questions

  1. What's your current conversion rate?
  2. How much traffic does this page get?
  3. What change are you considering and why?
  4. What's the smallest improvement worth detecting?
  5. What tools do you have for testing?
  6. Have you tested this area before?

Related Skills

  • cro: For generating test ideas based on CRO principles
  • analytics: For setting up test measurement
  • copywriting: For creating variant copy

04 同仓库更多 Skills

SanHsien
SanHsien/agent-marketing-skills

video

When the user wants to create video with AI tools or code, or asks which approach to use (screen recording, AI avatar, generated, or templated) or how to make videos repeatably: product demos, AI generation (Veo, Sora, HeyGen), programmatic video with Remotion. For what to post, see social. For video ads, see ad-creative.

★ 1
SanHsien
SanHsien/agent-marketing-skills

social

When the user wants organic social content for LinkedIn, X, Instagram, TikTok, or Facebook, or asks what to post, which format to favor, or from which account: posts, threads, carousels, short-form video scripts, calendars, build-in-public posts from commits, listening. No AI tells like 'it's not X, it's Y' or broetry. For content strategy, see content-strategy. For paid ads, see ad-creative. For press, see public-relations.

★ 1
SanHsien
SanHsien/agent-marketing-skills

sms

When the user wants SMS, MMS, or WhatsApp marketing, or asks whether texting is worth it or a message is allowed: abandoned cart, win-back, and promo texts, WhatsApp templates and windows, TCPA and A2P 10DLC compliance, SMS vs email. For email, see emails. For phone capture popups, see popups.

★ 1
SanHsien
SanHsien/agent-marketing-skills

site-architecture

When the user wants to plan or restructure a website, or asks which pages they need or where something belongs: hierarchy, navigation, URL structure, internal linking, website builders and migrations. Not XML sitemaps (see seo-audit). For structured data, see schema.

★ 1
SanHsien
SanHsien/agent-marketing-skills

signup

When the user wants better signup, registration, and trial start flows, or asks which fields or steps to cut or whether to require a card: SSO and drop-off before an account exists. For what happens after signup, see onboarding. For lead forms that don't create an account, see cro.

★ 1
SanHsien
SanHsien/agent-marketing-skills

seo-audit

When the user wants to audit or diagnose SEO, or asks why traffic dropped, a page won't rank, or whether a setup like canonicals, hreflang, or redirects is right: technical SEO, indexing, Google updates, on-page and titles, Core Web Vitals, local SEO. Start here when someone just says their SEO is bad. For pages at scale, see programmatic-seo. For structured data, see schema. For AI search, see ai-seo.

★ 1