⌘K
DARK
MENU
DARK

Last reviewed

Learn Experimentation and CRO: From Zero to Expert

Experimentation and CRO is the work of finding out what makes more visitors buy, sign up or enquire, then proving it with tests. This learning path takes you from zero to job-ready in four stages, with over 100 resources across videos, courses, guides and tools, and a project at every stage.

  • 161resources
  • 4stages
  • 6 to 9 monthsto job-ready
Start the roadmapSee the tools
Time to job-ready6 to 9 months at 5 to 10 hours a week
DifficultyMedium: statistics and judgement matter more than code
First job titleCRO Analyst or Experimentation Analyst
You needA laptop, GA4 access and a site to practise on
Learn firstGoogle Analytics 4
Cost to learnFree; paid testing tools are optional

What is experimentation and CRO?

Experimentation is the practice of testing a change on part of your audience before rolling it out to everyone. Conversion rate optimisation, or CRO, uses research, analytics and experiments to raise the share of visitors who take a useful action, such as buying, booking or signing up. A/B testing is its best-known method.

CRO sits between marketing, design, product and data. Marketing brings the traffic, design shapes the experience and analytics measures the result. CRO connects them and replaces opinions with evidence. It is also about learning: a test that loses still tells you something about your customers.

  1. 1MeasureSet up analytics so you can see how visitors move through the site and where they drop out.
  2. 2ResearchCombine behaviour data with interviews, surveys and usability tests to find out why.
  3. 3HypothesiseWrite a clear, testable idea and rank it against others by impact and effort.
  4. 4TestShow the original and a variant to random visitors and measure the difference.
  5. 5Learn and shipDecide, roll out the winner, record what you learned and feed it into the next idea.

What does an experimentation and CRO specialist do?

What a experimentation and cro specialist doesYour visitors goes in. Analyse funnels and behaviour, Run customer research, Write and rank hypotheses, Design and launch A/B tests, Analyse and report results, Build the testing habit. Tested and improved comes out, and results feed back into the next round.DATA AND RESULTS FEED THE NEXT ROUNDYOUR VISITORSTESTED ANDIMPROVEDEXPERIMENTATIONAND CRO1Analyse funnelsand behaviour2Run customerresearch3Write and rankhypotheses4Design andlaunch A/Btests5Analyse andreport results6Build thetesting habit
The work flows clockwise from 1 to 6. Each step is explained below.
  1. 1Analyse funnels and behaviourUse GA4 and heatmaps to find where visitors leave and which groups struggle.
  2. 2Run customer researchUse surveys, interviews and usability tests to find out why people hesitate.
  3. 3Write and rank hypothesesTurn findings into testable ideas and rank them with ICE, PIE or PXL.
  4. 4Design and launch A/B testsChoose the metric, calculate sample size, build variants and check tracking before launch.
  5. 5Analyse and report resultsRead results with the right statistics, then write clear recommendations.
  6. 6Build the testing habitKeep a backlog and a learning log, and help teams agree when to test and when to just ship.

How to learn experimentation and CRO: a step-by-step roadmap

Work through the stages in order. Each stage mixes formats so you can learn the way that suits you, then proves the skill with a checklist and a project. The hours are a rough guide for someone starting from zero. Use the filter to show only videos, only courses or only free resources.

Learning roadmap for Experimentation and CRO4 stages and 161 resources from beginner to job-ready.YOUR ROADMAPExperimentation and CRO4 stages, 161 resources, one project per stageBEGINNER10 TO 20Basics25 to 35 hours220 TO 50Core skills40 to 60 hours350 TO 80Practice50 to 80 hours480 TO 100Mastery60 to 100 hoursJob-ready: CROAnalyst orExperimentation
Your route through this learning path. Select a stage to jump to it.
Show
1

0 to 20About 25 to 35 hours

Stage 1: Experimentation and CRO basics

Understand what CRO is, read analytics and funnels in GA4, set up basic tracking and learn the statistics ideas behind every test.

What experimentation and CRO are

Learn what conversion rate optimisation is, how it differs from random website tweaking, and where A/B testing fits.

Key concepts
  • Conversion rate
  • Macro and micro conversions
  • Funnel
  • CRO
  • A/B test
  • Hypothesis
  • Control and variant

Web analytics and funnels in GA4

Read the numbers that show where visitors drop out. Every experiment idea starts here.

Key concepts
  • Events
  • Key events
  • Sessions and users
  • Funnel exploration
  • Segments
  • Traffic sources

Tags, events and a tracking plan

You can only test what you measure correctly. Set up tags and events, and write down what each one means.

Key concepts
  • Tag manager
  • Trigger
  • Data layer
  • UTM parameters
  • Event naming
  • Tracking plan

Statistics you need first

A/B testing is applied statistics. Learn samples, uncertainty and p-values before you touch a test.

Key concepts
  • Sample and population
  • Mean and variance
  • Central limit theorem
  • Confidence interval
  • Null hypothesis
  • P-value

Check yourself: you are ready for the next stage when you can

Project

Measure one funnel properly

Pick a site you can access, such as your own portfolio, a friend's shop or a demo store. Install GA4 and Tag Manager, write a short tracking plan for three events, build a funnel exploration for the main goal and write one page on where visitors drop out.

Deliver
A tracking plan, a funnel screenshot or export and a one-page summary of the biggest drop-off.
Done when
Events fire correctly, the plan matches what is tracked, and your summary names one drop-off with the numbers behind it.
2

20 to 50About 40 to 60 hours

Stage 2: Experimentation and CRO core skills

Find problems with research and behaviour data, write ranked hypotheses and design a test with a sensible metric and sample size.

User research and qualitative insight

Numbers show where people leave. Research shows why. Learn the methods that produce test ideas worth running.

Key concepts
  • Qualitative and quantitative
  • Interviews
  • Usability testing
  • Voice of customer
  • Heuristic evaluation
  • Research plan

Behaviour analytics and the CRO audit

Use heatmaps, recordings and funnels to find friction, then pull the findings into one audit.

Key concepts
  • Click map
  • Scroll map
  • Session recording
  • Rage clicks
  • Friction
  • CRO audit

UX and persuasion principles

Learn what makes a page easy to use and easy to trust, so your ideas rest on evidence, not taste.

Key concepts
  • Usability heuristics
  • Cognitive load
  • Social proof
  • Scarcity and urgency
  • Value proposition
  • Trust signals

Hypotheses and prioritisation

Turn research into clear hypotheses, then rank them with ICE, PIE or PXL so you test the right things first.

Key concepts
  • Hypothesis statement
  • ICE
  • PIE
  • PXL
  • LIFT model
  • Test backlog
  • Roadmap

A/B test design, sample size and significance

Plan a test that can give a trustworthy answer: the right metric, enough traffic and a clear stopping rule.

Key concepts
  • Primary metric
  • Guardrail metric
  • Sample size
  • Minimum detectable effect
  • Statistical power
  • Significance level
  • Test duration

Check yourself: you are ready for the next stage when you can

Project

Audit a real site and plan its tests

Audit a live site with Clarity, GA4 and a heuristic review. Write ten hypotheses, score them with a framework, then design the top three tests in full: primary metric, guardrails, sample size, duration and stopping rule.

Deliver
A CRO audit report and a prioritised test plan with three fully designed tests.
Done when
Each finding is tied to evidence, each test has a calculated sample size and a written decision rule, and a reader could run the tests without asking you questions.
3

50 to 80About 50 to 80 hours

Stage 3: Experimentation and CRO practice

Launch real tests, improve pages, forms, checkouts and pricing, read results fairly and keep your work ethical.

Building and launching tests

Set up a test in a real tool, check it on every device and launch it safely.

Key concepts
  • Variant
  • Traffic split
  • Visual editor
  • Feature flag
  • QA
  • Flicker
  • Goal tracking

Landing page and lead form optimisation

Improve the pages and forms where visitors decide to act.

Key concepts
  • Value proposition
  • Headline
  • Call to action
  • Above the fold
  • Form fields
  • Page speed
  • Message match

Checkout, ecommerce and pricing

Find and remove the friction between a ready buyer and a completed order, and test how prices are shown.

Key concepts
  • Cart abandonment
  • Checkout steps
  • Payment methods
  • Trust badges
  • Price anchoring
  • Plan comparison

Reading results and reporting

Decide what the data allows you to conclude, write it up, and keep what you learned.

Key concepts
  • Winner, loser and inconclusive
  • Confidence interval of lift
  • Segment analysis
  • Novelty effect
  • Learning log
  • Test report

Ethics, privacy and trustworthy design

Test to help people, not to trick them. Learn what deceptive patterns are, why consent matters and how to stay inclusive.

Key concepts
  • Deceptive patterns
  • Informed consent
  • Consent mode
  • Accessibility
  • Inclusive design
  • Honest social proof

Check yourself: you are ready for the next stage when you can

Project

Run and report a real test

Run an A/B test on a live page for a client, an employer, a volunteer organisation or your own site. If traffic is too low, run a simulated test on a demo store plus a moderated usability test, and say so clearly. Write the full report: hypothesis, set-up, results and next steps.

Deliver
A test report of two to four pages with screenshots of each variant.
Done when
The report states the hypothesis, sample size, dates, result with its range, decision and lessons, and another person has read it and found no gaps.
4

80 to 100About 60 to 100 hours

Stage 4: Experimentation and CRO mastery

Handle advanced statistics, run a programme, test across product and server-side and lead experimentation strategy.

Advanced experiment statistics

Learn why tests mislead experienced teams: sample ratio mismatch, peeking, sequential and Bayesian methods, variance reduction.

Key concepts
  • Sample ratio mismatch
  • Peeking
  • Sequential testing
  • CUPED
  • Bayesian testing
  • Multi-armed bandits
  • Twyman's law
  • OEC

Running an experimentation programme

Move from single tests to a team habit: velocity, governance, learning and culture.

Key concepts
  • Test velocity
  • Win rate
  • Governance
  • Maturity model
  • Learning repository
  • Centre of excellence

Product, server-side and feature-flag experimentation

Test product changes, pricing and algorithms in code, with the data team, and measure effects that last.

Key concepts
  • Server-side test
  • Feature flag
  • Warehouse-native
  • Holdout
  • Long-term effect
  • Incrementality
  • Personalisation

Strategy, leadership and the CRO career

Learn to win budget, work with stakeholders, shape a team and grow your career.

Key concepts
  • Business case
  • Revenue impact
  • Stakeholder management
  • Team structure
  • Portfolio
  • Specialisation

Check yourself: you are ready for the next stage when you can

Project

Build an experimentation programme case study

Choose a real organisation or your own project. Write a twelve-month experimentation strategy: maturity assessment, team and tools, governance, prioritisation scheme, reporting and a first-quarter roadmap of at least eight tests. Include one completed test analysis with sample size, sample ratio check and a result write-up. Publish it as a portfolio piece.

Deliver
A published case study with strategy, roadmap and one worked test analysis.
Done when
A hiring manager can see your reasoning, your numbers and your decisions, and every claim is backed by data you can show.

Experimentation and CRO tools to learn

Learn firstGoogle Analytics 4See how visitors move through your site, where they drop out and what converts.FreeLearn firstMicrosoft ClarityFree heatmaps and session recordings that show what visitors actually do.FreeGoogle Tag ManagerAdd tracking and events without waiting for a developer.FreeEvan's Awesome A/B ToolsSample size and significance calculators that many testers trust.FreeGrowthBookOpen-source platform for feature flags and A/B tests with built-in statistics.Free planPostHogProduct analytics, session replay and experiments in one tool.Free planVWOVisual editor testing, heatmaps and reporting; common in agencies.PaidOptimizelyEnterprise web and feature experimentation platform.Paid

Is experimentation and CRO a good career?

Yes, if you enjoy mixing data with human behaviour. Most businesses with a website or an app want more from the traffic they already have, and many have nobody who can prove what works. The skill also travels well: the same thinking applies to products, emails, pricing and ads.

It makes existing traffic worth more

Visitors from search and ads cost money. Raising the share who convert improves the return on all of it.

It replaces opinion with evidence

Tests settle arguments about design and copy with data, which makes you useful to every team.

It blends several skills

You use analytics, psychology, writing, design and statistics, so you can move into product, growth or data work.

Testing reaches beyond websites

Product, email and pricing teams run tests as well, which widens where you can work.

Experimentation and CRO career path and jobs

  1. Entry

    CRO Analyst or Experimentation Analyst

    Sets up tracking, builds reports, reviews heatmaps and helps build and check tests.

  2. Mid

    CRO Specialist or Optimisation Manager

    Owns the test backlog, designs experiments, analyses results and presents them.

  3. Senior

    Senior CRO Manager or Experimentation Strategist

    Shapes the roadmap with product and marketing, mentors analysts and sets statistical standards.

  4. Leadership

    Head of Experimentation or Head of Growth

    Owns the programme, budget and tools, and builds a testing culture across the company.

  5. Specialist

    Experimentation Scientist or Growth Engineer

    Builds testing platforms, designs advanced analyses and runs server-side tests with engineers.

  6. Independent

    CRO Consultant or Freelance Optimiser

    Audits sites and runs test programmes for several clients, or trains in-house teams.

Common experimentation and CRO mistakes beginners make

MistakeStopping a test the moment it looks significant.

Do thisSet the sample size and run time first. Stopping on a good day produces false winners.

MistakeTesting without enough traffic.

Do thisCalculate sample size before you start. If the page cannot reach it, test bolder changes or use research instead.

MistakeCopying another site's winning test.

Do thisTheir audience and page differ from yours. Treat other results as ideas, then test them yourself.

MistakeRunning tests without a hypothesis.

Do thisWrite down the observation, the change and the expected effect, so even a loss teaches you something.

MistakeIgnoring a sample ratio mismatch.

Do thisIf the visitor split differs from the plan, the result cannot be trusted. Check it before you read the lift.

MistakeWinning on one metric while harming another.

Do thisAdd guardrail metrics such as revenue per visitor, refunds and page speed, and check them before you ship.

How to get an experimentation and CRO job

Employers hiring a first CRO or experimentation analyst look for evidence that you can find a problem, design a fair test and explain the result clearly. Your portfolio matters more than certificates.

Job-ready checklist

  • You can build a funnel in GA4 and explain where people leave
  • You can write a hypothesis and rank it against other ideas
  • You can calculate sample size and explain the result of a test
  • You can review heatmaps and recordings and turn them into test ideas
  • You can spot a sample ratio mismatch and explain why it matters
  • You can talk through one real test from idea to decision

Portfolio pieces that get interviews

  1. A CRO audit of a real site, with a prioritised list of test ideas
  2. A full test report: hypothesis, sample size, result with its range and the decision
  3. A before and after case study with tracking evidence, even from a small project
  4. A short write-up of what you learned from a test that lost

Experimentation and CRO interview questions

A test shows a 12% lift after three days. Would you ship it?

A strong answer covers: Not yet. Check the planned sample size and run time, whether the split is correct and whether the result holds across weekdays and weekends. Early results swing widely. A good answer sticks to the plan agreed before launch.

How do you decide what to test first?

A strong answer covers: Start from evidence: analytics, research and past tests. Write hypotheses and rank them with ICE or PXL. Mention traffic limits, since the best idea is no use if the page cannot reach the sample size.

What is a sample ratio mismatch?

A strong answer covers: It is when the number of visitors in each variant differs from the planned split by more than chance allows. It points to a bug in assignment or tracking, so the results cannot be trusted until it is fixed.

Your test is inconclusive. What next?

A strong answer covers: Check the test had enough power and look at the range of plausible effects. Segment with care. Record what you learned, then test a bolder change or move to another problem. An inconclusive result is still information.

How would you test a pricing change responsibly?

A strong answer covers: Be clear about what is being tested, avoid misleading claims and watch guardrails such as refunds and complaints. Consider fairness between customers and the rules in each market, and get sign-off before testing prices.

Experimentation and CRO glossary

Conversion rate
The share of visitors who complete a chosen action.
Conversion rate optimisation
Improving that share through research and testing.
A/B test
A controlled experiment that compares two versions using random visitors.
Variant
A changed version of a page or feature in a test.
Control
The original version that a variant is compared against.
Hypothesis
A testable statement of what you will change, why, and what you expect.
Primary metric
The one measure that decides whether a test won.
Guardrail metric
A measure you watch to make sure a change does no harm.
Statistical significance
Evidence that a difference is unlikely to be down to chance alone.
Statistical power
The chance that a test detects a real effect of a given size.
Minimum detectable effect
The smallest change a test is designed to detect reliably.
Sample size
The number of visitors a test needs to reach a reliable result.
Peeking
Checking results early and stopping when they look good, which inflates false wins.
Sample ratio mismatch
A visitor split that differs from the plan, which signals a data problem.
CUPED
A method that uses pre-test data to reduce noise and shorten tests.
Bayesian testing
A way of reporting the probability that one version is better.
Multivariate test
A test of several element changes and their combinations at once.
Feature flag
A switch that turns a feature on for chosen users, used to run tests in code.
ICE
A scoring model that ranks ideas by impact, confidence and ease.

Experimentation and CRO FAQs

What is the difference between CRO and A/B testing?

A/B testing is one method. CRO is the wider practice of finding problems with analytics and research, forming ideas, testing them and learning from the results. You can do CRO without a test, for example by fixing an obvious usability fault, but testing is how you prove that a change helped.

Do I need to be good at maths?

You need to understand ideas rather than do hard sums. Learn averages, variation, confidence intervals, p-values and power, and use calculators for the arithmetic. What matters is knowing what a result does and does not tell you. Stage 1 covers the statistics you need with free videos.

How much traffic do I need to run A/B tests?

It depends on your baseline conversion rate and the smallest change you care about. Low-traffic sites need larger effects to reach a reliable result. Use a sample size calculator before you start. If you cannot reach the number in a few weeks, use research and bolder changes instead.

Do I need to code?

Not to start. Many tools have visual editors, and GA4 and Tag Manager need no code. Basic HTML, CSS and some JavaScript help you build variants and check tracking. Server-side and product testing need more engineering skill, so treat that as a later step.

How long does it take to learn experimentation and CRO?

With five to ten hours a week, most people finish the four stages in six to nine months if they do the projects. Real skill comes from running many tests on many pages, which takes longer. A small real test for a friend or local business speeds this up.

Which certificate should I get?

Certificates matter less than a portfolio. Google Skillshop offers free analytics training, and paid CRO courses exist if an employer funds them. Employers want to see real tests, clear reports and sound judgement. Start with the free resources here and spend money on tools only if you need them.

More learning paths in the growth marketing map