Mapping a competitor's full funnel is defined as walking their customer journey end to end from several entry points and recording every step, rather than analysing their site from outside. Almost all of it is published to anybody who asks. The ads sit in a public library, the landing pages are served to everyone, the abandoned-cart sequence arrives at any address you hand over, and the post-purchase half costs one small order. Five doors are worth walking: a paid ad, an organic result, an assistant answer, an abandoned cart, and a real purchase through to a return. The last two are where the findings are, because almost no competitor audit reaches them.
- Walk the funnel as a customer. Most of it is published; only their rates are private.
- Five entry points: paid ad, organic result, assistant answer, abandoned cart, real purchase.
- The abandoned-cart sequence shows what a competitor believes about why people hesitate.
- Buy something cheap once a year. It buys the whole post-purchase half nobody else audits.
- Record what each step asked, what it promised and how long it took. Save the opinions for the end.
Competitor research usually stops at the product page, which is roughly a third of the journey and the third they have optimised most. Last updated: October 2026.
Omniconvert has mapped store journeys across the CROBenchmark dataset of 7,000+ websites in 15+ industries, against 248+ audit criteria, over 13 years in eCommerce, and the part of a rival's funnel that explains their results is almost never the part a desk-based audit looks at. Nexus by Omniconvert is the AI for eCommerce growth engine that unifies commerce data and ranks the next action by True Profit, and it rests on the same premise as this method: the journey is a sequence of observable moments, not a single page. The scoring and cadence side of this work is in DTC competitor research stack.
What is actually visible from outside
It is worth separating the two categories properly, because teams assume far more is hidden than actually is.
Private: conversion rate, average order value, acquisition cost, retention, margin, spend. None of it is obtainable and all of it gets estimated by audits that should have left the cell blank.
Published: every ad they are running and for how long, since Meta operates its Ad Library as a public archive. Every landing page. Every product page, in every market they serve. Every lifecycle email, to any address that triggers it. Every word of their packaging, inserts and returns process, to anybody who places an order.
Read that second list again and notice how much of a funnel it covers. The reason competitor research feels data-poor is not that the information is hidden. It is that collecting it requires walking rather than reading, and walking takes three weeks of calendar time.
How to map a competitor's full funnel: the five doors
Walk them in this order, because each one tells you what to look for in the next.
- Enter through a paid ad. Find a live ad in the public library, follow it, and record whether the landing page repeats the ad's promise in its first screen or starts the argument from the beginning. The gap between those two is a real competence difference.
- Enter through organic search. Search a buying question in your category and arrive on whatever they rank with. Note the page type and how many clicks it sits from a purchasable product.
- Enter through an assistant. Ask the same buying question of two or three assistants. Record whether they are named, what is said about them, and which page the answer points at.
- Abandon a cart and wait. Add to cart, give a fresh email address, leave. Then record every message: when it arrived, what it led with, and how many followed.
- Buy something cheap and keep everything. One small order. Capture the confirmation, the dispatch note, the packaging, the inserts, the delivery date against the promise, the review request. Then return it and capture that too.
Door five is the one that gets cut for time and it is the one that cannot be obtained any other way. Budget forty pounds and three weeks per competitor and it stops being a decision you have to argue for each quarter.
What to capture at each step, and what to ignore
The discipline here is subtraction. A walk that records everything produces a document nobody reads.
What it asked of you is the friction measure: an email, an account, a phone number, a choice between four variants. Count the asks rather than describing them.
What it promised is the message measure: delivery by a date, a return window, free shipping over a threshold, a guarantee. Record the exact words, because the precision of a promise is itself a finding.
How long it took is the only quantitative field you can reliably collect, and it is the one most audits omit. Time from click to page interactive. Days from order to dispatch. Days from dispatch to arrival against the stated promise. Hours from return request to label.
The table sets out the five doors with what to capture and the finding each one typically produces.
| Door | What to capture | Typical finding |
|---|---|---|
| Paid ad | Ad promise, landing page first screen | Whether message and page match |
| Organic result | Page type, clicks to a product | How far content sits from purchase |
| Assistant answer | Named or not, and the page sent | Which facts they publish readably |
| Abandoned cart | Timing, first offer, message count | What they think causes hesitation |
| Real purchase | Promise against actual, inserts, review ask | Whether operations match marketing |
| Return | Steps, who pays, hours to resolve | The gap between policy and practice |
The last two rows are the half of the funnel that a desk audit never sees, and they are where a competitor's actual operating quality lives. A store with excellent pages and a six-day dispatch time is a different competitor from the one its website describes.
Reading the abandoned-cart sequence
Of the five doors this is the cheapest and the most revealing, so it deserves its own reading.
Note the first message's timing. An hour means they believe the decision is still live. Twenty-four hours means they think it is a considered purchase. Three days means nobody has looked at the sequence in two years.
Note what it leads with. A discount in the first message is an expensive default and a strong signal that they have not diagnosed their abandonment. Reassurance, a delivery promise, a returns reminder or a review quote all suggest somebody looked at why people were leaving. Baymard Institute's research on abandonment reasons is the obvious external check on whether their guess is a good one.
Note how many messages follow and whether the offer escalates. An escalating discount across three messages teaches a customer to abandon deliberately, which is a weakness you can decline to copy.
Then do the same walk on your own store with a fresh address, which almost nobody does, and compare the two sequences side by side. The comparison is frequently uncomfortable and always useful.
What a growth team should do this week
Start with one competitor rather than three, because the method is what you are learning first.
- Walk the three digital doors on your closest rival. Paid ad, organic result, assistant answer. One afternoon, screenshots with dates.
- Trigger the abandoned cart today. A fresh address, a real product, then leave. The waiting is the slow part, so start it before anything else.
- Place one small order and diarise capturing every step through to a return. Forty pounds buys the half of the funnel nobody audits.
- Walk your own funnel identically with a fresh address and no internal shortcuts. This is the comparison that produces the decision.
- Write the summary once, after all five walks, in one page: what they ask, what they promise, how long they take, and where you are slower.
Where ad data, journey observations and your own analytics have to sit together before a gap can be sized against revenue, that unification is the problem Nexus addresses, with every proposed change staying something a person approves before it goes live. If you would rather be scored against real stores in your category and country than against one rival you walked, Ecommerce Benchmark places a store on the free leaderboard across six dimensions, including the AI Visibility and Agentic Commerce axes that a manual walk can only sample.
FAQ: mapping a competitor's funnel
How do you map a competitor's funnel without their data?
By walking it as a customer rather than analysing it as an outsider. Their funnel is published: the ads are in a public library, the landing pages are served to anybody, the abandoned-cart sequence arrives at any address you give it, and the post-purchase experience costs the price of one small order. What you cannot get is their rates. What you can get is every step and every message, which is most of what you wanted.
Which entry point reveals the most?
The abandoned cart, because it shows what a competitor believes about why people hesitate. The timing of the first message, whether it leads with a discount or with reassurance, and how many messages follow all encode a decision somebody made about their own funnel. It is also the cheapest of the five entry points to walk, requiring only an email address and some patience.
Is it worth buying from a competitor?
Yes, once a year per competitor, and it is the step most audits skip. One small order buys the confirmation sequence, the packaging, the inserts, the delivery timing against the promise, the review request and, if you return it, the whole returns experience. That is half the customer journey and none of it is visible from outside the purchase.
What should you record at each step?
Three things and no more: what the step asked of you, what it promised, and how long it took. Screenshots of everything, with the date in the filename. Resist writing an opinion at the time, because the judgement is better made later across all five walks, when a pattern is visible that no single step showed you.
How long does mapping one competitor's funnel take?
About three hours of active work spread over three weeks, because the waiting is unavoidable. The four digital doors take an afternoon. The purchase walk takes as long as their delivery and returns cycle, which is itself one of the findings. Plan it as a calendar exercise rather than a sprint, and run one competitor at a time.
The bottom line
Stop auditing competitors from your desk and go and be their customer. Walk in through a paid ad, an organic result and an assistant answer in one afternoon, then trigger their abandoned-cart sequence and let it run while you wait. Place one small order and record the promise against what actually happened, then return it and record that too. Three fields per step, what it asked, what it promised, how long it took, and screenshots with dates. Then do the identical walk on your own store, which is the part that makes the whole exercise land, because the gap you can act on is almost always in the half of the funnel neither of you has been looking at.