How to use Path Analysis
Path analysis maps the actual sequences shoppers take through your store, surfacing patterns funnel reports miss. Here's how to run it without drowning in noise.
Path Analysis
The quantitative study of the page sequences users follow before converting, abandoning, or churning.
Path analysis looks at the ordered series of pages a visitor touches in a session — product → reviews → size guide → cart → checkout — and counts how often each sequence appears among converters versus non-converters. The output is a ranked list of high-value and high-risk routes through your store.
Where a funnel report tells you that 38% of shoppers drop between cart and checkout, path analysis tells you which detours they took on the way out — a return policy page, a shipping FAQ, a size guide they couldn't escape. It turns navigation behaviour into testable hypotheses.
Most teams already have funnel reports. They tell you that conversion drops 42% between product page and add-to-cart, which is useful — until the question becomes why. Path analysis is the layer underneath: the actual page-by-page itinerary every session takes.
It works best when paired with funnel analytics rather than replacing it. Funnels frame the problem; paths reveal the mechanism. A 6% sitewide conversion rate hides a story where shoppers who view the FAQ convert at 14% and shoppers who get stuck on the size guide convert at 1.9%.
What path analysis actually surfaces
The headline output is a list of frequent sequences ranked by conversion rate. On a Shopify apparel store you'd typically see five to twelve dominant paths covering 60-70% of all sessions — the rest is a long tail of one-off journeys.
The interesting signal is rarely the top path. It's the unexpected ones: the converters who all viewed a specific blog post, the cart abandoners who all loaded the shipping calculator twice, the returning customers who skip the homepage entirely and deep-link straight to a collection.
Three patterns show up on almost every audit. First, a confidence-building page (reviews, FAQ, size guide done right) that converters disproportionately visit. Second, a friction page where non-converters loop. Third, a backward step — checkout → shipping policy → never returns — that signals a missing answer.
Forward paths vs reverse paths
Forward path analysis starts from an entry page and asks 'where do people go from here?'. Reverse path analysis starts from a goal (a thank-you page, a cart) and asks 'how did they get here?'. Reverse is usually the higher-leverage view for CRO — you're working backward from outcomes you already know matter.
How to run a path analysis without drowning in noise
The single biggest mistake is analysing all traffic in one pool. Paid social visitors behave nothing like email-list returners, and mixing them produces an average that describes nobody. Always segment by channel, device, and new-vs-returning before you look at any path.
Then cap path length. Sessions can include dozens of pageviews, but anything past the fifth or sixth step is mostly noise — browsers tab-hopping or comparison shopping. Most actionable patterns live in 3-5 step sequences.
Conversion rate by path length (apparel, mobile, paid social)
The curve above is typical: conversion peaks at 4-5 pages and decays sharply after that. Long sessions usually mean a shopper couldn't find what they needed, not that they were highly engaged. Treat 7+ page sessions as a research signal, not an engagement win.
Benchmarks for common path patterns
The table below compares conversion rates for the path archetypes we see most often on Shopify and WooCommerce stores in the €1M-€15M range. Use it as a directional reference — your own numbers will vary with product, price point, and channel mix.
What matters is the gap between archetypes on your store, not the absolute number. If your 'product → reviews → cart' path converts 3x better than 'product → size guide → cart', you have a size-guide problem worth fixing this sprint.
Typical conversion rate by path archetype, by vertical
| Path archetype | Apparel | Beauty | Electronics |
|---|---|---|---|
| Landing → PDP → Cart → Checkout | 3.8% | 4.4% | 2.9% |
| Landing → PDP → Reviews → Cart → Checkout | 6.1% | 7.2% | 5.4% |
| Landing → Collection → PDP → Cart → Checkout | 4.2% | 4.0% | 3.1% |
| Landing → PDP → Size guide → PDP → Cart | 1.9% | — | — |
| Landing → PDP → Shipping FAQ → Exit | 0.4% | 0.6% | 0.5% |
| Email → PDP → Cart → Checkout (returning) | 9.8% | 11.3% | 7.6% |
Two rows in that table earn their keep on every audit. The size-guide loop on apparel — visitors who go PDP → size guide → back to PDP convert at less than half the baseline. And the shipping-FAQ exit, which is almost always a missing piece of information that belongs on the product page itself.
Common traps and how to avoid them
Path analysis is correlational. A page that appears in many converter paths isn't necessarily causing those conversions — high-intent shoppers might just visit more pages on the way to a purchase they were going to make anyway. The only way to confirm causation is to A/B test removing or surfacing the page.
Watch out for survivorship in your data. If you only analyse sessions that reached step three, you've already filtered out the dropouts whose behaviour is the most informative. Always include the abandoners and tag their last-seen page — that's where the real friction lives.
Don't act on paths with fewer than 200 sessions
A 'high-converting path' built on 40 sessions and three buyers will mislead you. Set a minimum sample size before any path gets escalated to a hypothesis — 200 sessions per segment is a reasonable floor, 500+ before you commit dev time to a change.
Path analysis FAQ
A funnel measures progression through a predefined sequence of steps you specify in advance — view product, add to cart, checkout, purchase. Path analysis is open-ended: it discovers the sequences shoppers actually take, including the detours and loops you didn't anticipate. Funnels answer 'how many drop at each step?'; paths answer 'what did they do instead?'.
Yes, GA4 has a Path Exploration report under Explore. It's serviceable for simple forward and reverse paths but gets unwieldy past three or four steps and doesn't segment by conversion outcome cleanly. Most CRO teams export the data or use a dedicated tool once they're past basic exploration.
Around 5,000 sessions per month is a practical floor for sitewide patterns, and at least 200 sessions per segment-path combination before you act on anything. Below that, the long tail of unique paths swamps any real signal and you'll chase patterns that don't replicate.
Include them as a single-page path category, but analyse them separately. Bounces tell you about landing-page-message-match — a different problem from on-site navigation. Mixing them into multi-step path analysis drags every average down and obscures the patterns that matter for engaged shoppers.
Most stores use 28 or 30 days to smooth out weekly seasonality. Go shorter (7-14 days) only if traffic is high enough to give you 500+ sessions per segment in that window. Avoid windows that span a major promotion or holiday — the path mix shifts so much it muddles the baseline.
Only if you have user-level identity stitching (logged-in users, email matching, or a stable identifier). Otherwise a shopper who browses on mobile and buys on desktop looks like two separate sessions with two broken paths. Segment by single-device sessions if you don't have stitching — the data is cleaner.
A loop is when a shopper returns to the same page two or more times in a session, often PDP → size guide → PDP or PDP → reviews → PDP. Reviews loops usually correlate with higher conversion (confidence building). Size guide and shipping loops usually signal missing information — the shopper keeps checking because their question isn't answered.
Pick a path with a clear hypothesis: 'shoppers who detour to the shipping FAQ convert at 0.4% vs 3.8% baseline.' The test is to surface shipping cost on the PDP itself and measure whether the detour rate drops and overall conversion rises. The path data gives you the leak; the test confirms whether closing it actually moves revenue.
Yes, and it's often the highest-leverage segment. Returning customers usually have radically shorter, more direct paths — email → PDP → checkout — and any friction shows up dramatically because they're not browsing. If returners are taking longer paths than they used to, something has changed in your navigation or merchandising.
Critically so. Paid traffic is the most expensive on your site, so any wasted detour costs real money. Path analysis on a paid-social segment often reveals that landing-page → exit accounts for 50%+ of sessions — meaning your ad-creative-to-landing-page match is the leak, not your checkout flow.
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.