How to use User Journey Mapping
A practical guide to mapping multi-session user journeys across channels and touchpoints — what to capture, how to read it, and where it changes your CRO roadmap.
User Journey Mapping
Visualising the full multi-session path a shopper takes from first touch to purchase across channels, devices, and time.
User journey mapping is the practice of reconstructing the real sequence of sessions, channels, and pages a shopper moves through before they convert — or before they drop off. Unlike a funnel report, which collapses behaviour into stages, a journey map keeps the order, the gaps, and the channel switches intact.
For an online store, a typical map covers paid acquisition, organic and direct returns, email and SMS re-engagement, product and category browsing, and the checkout itself. It is a strategic artifact: marketing uses it to decide where to spend; UX and CRO teams use it to decide what to test next.
Most teams already have the raw material — GA4 events, server logs, email opens, ad clicks — but it sits in five tools that do not share an identity. Journey mapping is the work of stitching those touchpoints back into a single human story so the patterns become legible.
The output is not a poster on the wall. A useful journey map is a queryable dataset: every user, every session, every touchpoint in order. From there you can answer questions like 'what does the second visit look like for buyers versus non-buyers?' — which is where the testable hypotheses come from.
What a journey map actually contains
At minimum you need five fields per touchpoint: user identifier, timestamp, channel, landing page or event, and device. With those you can already reconstruct the sequence and measure gaps between visits.
More mature maps add product-level interest signals (which collection or PDP they viewed), engagement depth (scroll, video, size-guide opens), and any explicit intent moments — added to cart, started checkout, used a discount code. This is where journey mapping overlaps with broader behavioral analytics: the events are the same, but the analytical lens is sequential rather than aggregate.
The hardest field is identity. A shopper who clicks a Meta ad on mobile, returns the next day via Google search on desktop, and finally converts through a Klaviyo email is one person to your business and three users to GA4. Server-side tracking, hashed-email login events, and post-purchase reconciliation are the usual fixes.
Last-click attribution will lie to you here
If you only look at the converting session, 60-80% of journeys appear to come from 'direct' or 'branded search' — because that is the final touch, not the journey. Map the full path before you defund a channel.
How the sessions actually distribute
Most teams assume the journey is shorter than it is. For considered apparel and beauty purchases between €40 and €150, the median buyer takes three to five sessions over six to nine days. Sub-€30 impulse items collapse to one or two sessions; electronics and home goods stretch to seven-plus.
The chart below shows the channel mix at each session number for a representative mid-AOV apparel store. The pattern repeats across stores in this band: paid acquisition dominates session one, organic and direct take over by session three, and email closes the deal.
Channel share by session number (mid-AOV apparel store)
Paid social/search
Organic/direct
Email/SMS
Referral/other
Read this as a brief, not a verdict. It tells you where to look — for example, the gap between session one and session two is where your retargeting and welcome flow are doing their job, or aren't. If session-two recovery sits below 20% of session-one visitors, the map has just handed you a project.
Benchmark journey shapes by vertical
Journey length and shape vary more by vertical and average order value than by platform. Beauty buyers in the €20-€50 band convert quickly off a discount-triggered email; consumer electronics buyers spend a week reading reviews and comparison content before they even land on your PDP.
Use the table below to sanity-check your own numbers. If your apparel store is converting in 1.4 sessions on average, you are either reading the data wrong (identity stitching is broken) or you are leaving repeat-visit revenue on the table.
Typical journey shape by vertical (mid-AOV online stores)
| Vertical | Median sessions to purchase | Median days to purchase | Touchpoints per buyer | Top conversion channel |
|---|---|---|---|---|
| Beauty & personal care | 2-3 | 3-5 | 5-7 | Email/SMS |
| Apparel & accessories | 3-5 | 6-9 | 7-10 | Organic/direct |
| Home & lifestyle | 4-6 | 8-14 | 9-12 | Organic search |
| Consumer electronics | 5-8 | 10-21 | 12-18 | Comparison/review referral |
| Food & supplements (subscription) | 1-2 | 1-3 | 3-5 | Paid social |
The 'touchpoints per buyer' column matters most for media planning. If you need twelve touches to close an electronics sale, a campaign optimised on seven-day click attribution will under-credit the upper-funnel work that made the close possible.
Turning the map into a CRO roadmap
A journey map earns its keep when it changes the next test you run. Start by segmenting buyers from non-buyers and looking only at where the paths diverge — that is your hypothesis surface. Common divergence points are PDP scroll depth, size-guide opens, and whether session two starts on a category page or a product page.
From divergence points you can write concrete experiments: a sticky size-guide for the apparel buyers who never opened the modal, a returning-visitor PDP variant that surfaces reviews above the fold, a cart abandonment email triggered at the median gap rather than the default 24 hours.
The day-one shortcut
If you have at least six months of GA4 data, you can map your existing journeys without instrumenting anything new. Historical session, channel, and event data is enough to find the top three drop-off seams — and to brief your first three tests this week.
Frequently asked questions
A funnel collapses behaviour into ordered stages and counts users at each one. A journey map keeps the actual sequence, channel, and time gap of every touchpoint per user. Funnels tell you where people drop; journey maps tell you what they were doing before they dropped.
Behavioral analytics is the broader discipline of analysing what users do on your site — clicks, scrolls, rage taps, session recordings. Journey mapping is a specific sequential lens within it: same events, but stitched across sessions and channels into a path.
For a mid-AOV store you want at least 1,000 completed purchases in the window you're analysing, which usually means three to six months of data. Below that, journey patterns are noisy and you'll over-fit to a handful of unusual buyers.
Yes. Shopify exposes customer, order, and event data through its native APIs, and most analytics tools can stitch GA4 client IDs to Shopify customer IDs at checkout. The harder part is cross-device stitching before login, which needs server-side tracking or a tool that does it for you.
The standard pattern is to hash the email at the first known event (newsletter signup, login, checkout) and back-fill that identity onto all prior anonymous sessions sharing the same client ID. Anything before the first login on a new device stays anonymous unless you use deterministic cross-device tools.
Map both. Converting journeys tell you what works; non-converting journeys with high engagement depth tell you what almost worked — and those are usually the most actionable for CRO. The comparison between the two is where most insight lives.
Quarterly for the strategic view, plus an ad-hoc refresh whenever a major channel mix change happens — new paid campaign, iOS tracking shift, a promotion period. The shape of journeys changes faster than most teams expect.
Start with the five fields named above (user, timestamp, channel, landing page, device) and add product- and engagement-level events as you find specific questions to answer. Over-instrumenting up front creates noise; you can always backfill from existing event streams.
It informs attribution but doesn't replace it. A journey map shows you what really happens; an attribution model assigns conversion credit. Most teams use the journey map to choose a sensible attribution model (e.g. position-based for long electronics journeys, last non-direct for short impulse ones).
Find the largest single drop-off between two sequential touchpoints among engaged users — usually session-one PDP to session-two return, or cart to checkout. Design a variant that addresses the most likely cause (price anxiety, missing info, friction) and test it. That's almost always your highest-impact first move.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.