MarketingBeginnerPreview
Conversion Rate Optimization
A practical, evidence-driven course that teaches you to find where visitors leak out of your funnel, form testable hypotheses, run statistically valid experiments, and turn more of your existing traffic into customers.
Beginners, marketers, founders, and product people who own a website or funnel and want measurable conversion gains without a data-science background.
Course content
Workbook & downloads
Put the course into practice — a printable workbook plus editable templates you can fill in and reuse.
Preview the workbook
This workbook turns the course into reps you can run on a real page. Each section mirrors one course module with hands-on exercises, fill-in worksheets, and checklists you apply to your own funnel. Pick one website, landing page, or funnel you own and carry it through every section, and you will finish with a measured funnel, a researched conversion audit, a prioritised test backlog, and your first valid A/B test plan with the sample-size math already done.
What CRO Really Is and How to Think About It
Build an accurate model of conversion optimization — the funnel maths, the friction-and-motivation lens, and the loop — before you change anything.
Worksheet: Baseline Your Conversion Numbers
Pull your current numbers from analytics and compute the metrics that matter. Do this per segment, not just site-wide, so the averages stop hiding the truth.
- Macro conversion goal (purchase / paid signup / booked call) and its current overall rate
- Conversion rate split by device (desktop vs mobile)
- Conversion rate split by traffic source (e.g. paid social vs branded search)
- Conversion rate split by new vs returning visitors
- Average order value and revenue per visitor (revenue / visitors)
Exercise: Map Your Funnel and Find the Biggest Leak
List the ordered steps a visitor takes toward your goal and the number (or percentage) reaching each step. The single steepest drop is your headline opportunity for the rest of the workbook.
- Write each funnel step in order (e.g. land, view product, add to cart, begin checkout, purchase) with its count.
- Calculate the percentage drop-off between each consecutive step.
- Circle the step with the largest percentage drop — this is your primary target.
- For that step, label the likely bottleneck using the Fogg lens: low motivation, low ability (friction), or missing prompt.
Checklist: CRO Mindset Gut Check
- I am targeting the leakiest funnel step, not a random page or element.
- I am tracking revenue per visitor, not just conversion rate, so a vanity win cannot fool me.
- I have looked at conversion rate by device, source, and visitor type — not only the blended average.
- I am looking to remove friction before trying to add persuasion.
- I have framed my target leak as a motivation, ability, or prompt problem.
Research: Finding Where and Why Visitors Leave
Combine the quantitative 'where' from analytics with the qualitative 'why' from recordings and polls, then synthesise it into a ranked audit.
Worksheet: GA4 Funnel Exploration Findings
Build a Funnel exploration in GA4 for your real journey and record what it shows. Verify your key events fire correctly in DebugView first, or every number below is suspect.
- Key events defined and verified in DebugView (list them)
- Funnel steps as configured, in order
- Abandonment percentage at each step
- The single step + segment (e.g. mobile begin_checkout) with the biggest leak
- Page speed of that step from PageSpeed Insights (mobile score)
Exercise: Watch Ten Session Recordings
Open Microsoft Clarity (free) or Hotjar on your leakiest step. Watch at least ten recordings and study the heatmaps. Note real behaviour, not assumptions.
- Scroll heatmap: what important content sits below where most visitors stop scrolling?
- Click heatmap: where are people clicking that is not a link (dead clicks), and which buttons are ignored?
- Rage clicks / frustration signals: where did Clarity flag repeated frantic clicking?
- Recordings: describe two specific moments of hesitation, looping, or abandonment you watched.
Worksheet: On-Site Poll Setup and Answers
Launch one on-site or exit-intent poll on the target page and collect responses in the visitor's own words. These answers become both hypotheses and test copy.
- Page and trigger (e.g. exit-intent on checkout, on-load on pricing)
- The single poll question asked (e.g. what stopped you from completing your purchase today?)
- Top 3 recurring answers, verbatim
- The most common objection these answers reveal
- A candidate copy or design change the answers suggest
Checklist: Conversion Audit Quality Check
- Every key page is scored low / medium / high on all six heuristics (value prop, relevance, clarity, friction, distraction, urgency & proof).
- Each finding is written as a problem plus the evidence, not as a pet solution.
- Each finding is tagged with the funnel step and segment it affects.
- Each finding is labelled friction, motivation, or prompt.
- Findings are ranked so the highest-impact problems sit at the top.
From Insight to Hypothesis to a Ranked Backlog
Turn audit findings into falsifiable if-then-because hypotheses, then score and sequence them so you always test the highest-leverage idea first.
Worksheet: Write Three If-Then-Because Hypotheses
Take your top three audit findings and convert each into a falsifiable hypothesis. Use the LIFT Model to name which force you are strengthening or removing. Every hypothesis must be able to turn out false.
- Hypothesis 1 — If [change], then [metric + direction], because [evidence]. LIFT force affected:
- Hypothesis 2 — If [change], then [metric + direction], because [evidence]. LIFT force affected:
- Hypothesis 3 — If [change], then [metric + direction], because [evidence]. LIFT force affected:
- For each, confirm the failure case: what result would prove the hypothesis wrong?
Exercise: Score Your Backlog With ICE and PXL
Score at least five test ideas with ICE for a quick rank, then re-score your top two using PXL's concrete yes/no questions to pressure-test them. Note where the two methods disagree.
- ICE: score five ideas on Impact, Confidence, and Ease (1-10 each) and sort by total.
- PXL: for your top two, answer the factual questions (above the fold? high-traffic page? noticeable in 5 seconds? backed by user research? backed by a recording? effort to build?).
- Where did PXL change the ranking versus ICE, and why?
- Name the single highest-priority test you will run first.
Worksheet: Sequence Your Roadmap
Plan the order of your next few tests to mix quick wins that build credibility with at least one bold structural bet that could produce a large gain.
- Quick win #1 (high confidence, low effort) and the metric it targets
- Quick win #2 and the metric it targets
- One bold structural test (value proposition or page layout) and its hypothesis
- How you will re-score related ideas after each result
Checklist: Hypothesis and Prioritisation Audit
- Every test idea is written as an if-then-because hypothesis tied to research evidence.
- Each hypothesis names a metric and a direction and could be proven false.
- Every idea has an ICE or PXL score and a resulting rank.
- I am testing my highest-leverage idea first, not the easiest or my favourite.
- Every idea is logged in one backlog with page, step, segment, and status.
Running Valid Experiments and Reading Results
Run an A/B test that does not lie to you — compute the sample size, avoid the peeking trap, then read the result with confidence intervals and segments.
Worksheet: Calculate Sample Size and Runtime Before You Launch
Use a free calculator (Optimizely, VWO, or Evan Miller) to size your test BEFORE launching. Decide the stopping rule now so you cannot fool yourself later.
- Baseline conversion rate of the target metric (e.g. 3%)
- Minimum detectable effect you care about (e.g. 10% relative lift)
- Confidence level (aim 95%) and statistical power (aim 80%)
- Visitors needed per variation (from the calculator)
- Estimated runtime = visitors needed / daily traffic to this page (round up to full weeks)
Worksheet: Design One Clean A/B Test
Specify a single, clean test of your top-priority hypothesis. Change only one thing so you know what caused any difference, and pick the testing tool you will use.
- Hypothesis under test (if-then-because)
- Control (A) and Variation (B) described — the single element that differs
- Testing tool (VWO / Convert / Optimizely / PostHog / GrowthBook) and anti-flicker snippet installed?
- Primary conversion metric (the one you sized the test on)
- Pre-committed end date and the sample size you will not stop short of
Exercise: Interpret a Finished Result Correctly
When the test reaches its planned sample and runtime, read it properly. Go beyond the headline lift to the confidence interval and the segments before you decide what to ship.
- Did the test reach its planned sample size AND full runtime (at least 1-2 weeks)? If not, keep running.
- Record the reported lift and its confidence interval — does the interval include zero or a loss?
- Break the result down by device, source, and new vs returning — any segment winning or losing differently?
- Decide and justify: ship to all, ship to a winning segment, iterate, or discard.
Checklist: Valid-Test and Learning Check
- I set the sample size and end date before launch and did not call a winner early (no peeking).
- Only one element changed between A and B.
- The test ran at least one to two full weeks to cover weekday and weekend behaviour.
- I read the confidence interval, not just the headline percentage.
- I wrote a one-line insight in the backlog — whether the test won, lost, or was flat.
Your Action Plan
- Choose one website, landing page, or funnel you own to carry through the whole plan.
- Baseline your conversion rate, average order value, and revenue per visitor — split by device, source, and new vs returning.
- Map your funnel, calculate the drop-off at each step, and mark the single steepest leak as your target.
- Verify your GA4 key events in DebugView, build a Funnel exploration, and confirm the leakiest step and segment.
- Watch ten session recordings and the heatmaps on that step, and run one on-site or exit-intent poll.
- Synthesise everything into a conversion audit, scoring each key page on the six heuristics and writing findings as problems plus evidence.
- Convert your top findings into if-then-because hypotheses, naming the LIFT force each one affects.
- Score your ideas with ICE, pressure-test the top two with PXL, and log them all in one prioritised backlog.
- Calculate the required sample size and runtime for your top test, and write down the stopping rule before launch.
- Run one clean A/B test (one variable, anti-flicker installed) to full sample and runtime, then read the confidence interval and segments before shipping, and log the insight either way.
Pairs well with
Courses members commonly take alongside this one.
Flagship CoursePreview
Freelance Business Foundations: Position, Price, Sell, and Deliver High-Value Services
Freelancing · Beginner · 16h
Self-pacedPreview
Client GrowthPreview
Freelance Client Acquisition: Outreach, Leads, Referrals, and Deal Flow
Freelancing · Beginner · 15h 30m
Self-pacedPreview
Sales SystemPreview
Freelance Sales & Proposals: Discovery Calls, Scoping, Objections, and Closing
Freelancing · Intermediate · 16h
Self-pacedPreview