Find the website drop-off
Turn one unclear conversion path into a measurable investigation.
Start this Play →A conversion-event definition, sample funnel, investigation notes, and one change proposal.
Builders who have a website and want to understand one visitor journey before changing it.
- Estimated effort
- 30–60 min sample exercise
- Cost / access
- Free sample / usage costs vary
- First move
- Write the one action that counts as a successful visit, then check whether it is actually tracked.
Prerequisites, inputs & cost details
Before you begin
- A browser and the sample below; no account is needed for the exercise.
- For a real investigation: authorized analytics access and confirmed tracking/privacy settings.
Inputs you need
- A clear user journey and conversion event.
- Authorized event counts with a consistent date range; sample rows work here.
The sample is text-only. Check PostHog’s current included usage, analytics/replay costs, and data settings before installing.
Make it your next small win.
My saved workWrite a brief, practice the workflow, and keep your notes. This workspace saves your planning; it does not run tools or send anything.
See the fictional sample
Fictional aggregate events only. No visitor identifiers, tracking, or session recordings.
step,users,period Landing view,120,Sample week Pricing view,54,Sample week Signup started,18,Sample week Signup completed,9,Sample week Tracking note,unknown,Signup event needs verification
Investigate a fictional landing-to-signup journey. First verify that signup_completed means a successful account creation and that all counts cover the same eligible visitors and date window. Mark the sample funnel’s largest proportional drop, list two possible explanations, and propose one change with an owner and a measurement plan. Do not claim a cause or predicted uplift from aggregate counts.
- Define the conversionWrite the intended user action, exact event, eligible visitors, and time window. Check how the event is emitted.
- Build the sample funnelCompare the fictional steps with the same window and audience. Calculate counts and proportions, keeping untracked activity unknown.
- Investigate two explanationsCheck the experience manually first. In a real project use authorized, masked replay only after reviewing consent and data settings.
- Propose one changeChoose one small change with an owner, event checks, baseline, observation window, and rollback point.
- Set the reviewDescribe what you would compare after approval and deployment; include sample size and outside changes as limitations.
Review before a real build.
- A human reviews event definitions, privacy settings, and any proposed site change before implementation.
- Use only authorized analytics; masked replay is optional and must exclude private inputs.
- No analytics are installed, no recordings opened, and no website changes published by this exercise.
Measures to consider
- Share of eligible visitors completing the explicitly defined conversion.
- Verified tracking coverage and count of unresolved measurement questions.
Primary source checked · Source review 2026-11-01. PostHog documents conversion goals and event-based conversion measurement. This sample is fictional, not measured website performance or evidence that a particular change will improve conversions.
Understand the approach.
Start with a clear question: where does the intended next action stop happening?
01What you will make
+
- A narrow investigation into one conversion path, with event definitions and a reviewable change proposal.
- Aggregate numbers suggest where to investigate; they do not establish why a visitor left.
02Use it on a real website
+
- Confirm authorization, current tracking, privacy requirements, and whether the native analytics already answer the question.
- Compare like-for-like audiences and periods. Review event quality before considering a redesign.
- Keep replays optional, masked, and restricted to approved access; never include sensitive visitor input in your brief.
- Ship and measure a change only after the owner approves it.
Tools and supporting resources
Source recommendations; check current access, costs, API availability, and setup in the linked documentation. No integration is installed by opening this guide.
Plan your version.
Read the planning prompt
Help me investigate one website conversion path using the fictional sample or data I am authorized to share. Define the success event, verify tracking assumptions, create a small funnel, and separate observations from hypotheses. Propose one measurable change with human review, costs, privacy checks, and a rollback plan. Do not install tracking, access replay, spend money, change the site, or publish anything until I approve.