Customer journey mapping: the map that comes from data

You can draw a customer journey map in an afternoon. That is exactly what is wrong with it.

What comes out looks like a measurement and is a guess. Five boxes left to right, arrows between them, channels along the bottom. Everyone in the room agrees, because everyone is carrying the same story: people find us, like us, get in touch.

That buying decisions do not run that way has been researched since 2020 — the piece on the marketing funnel covers that finding and where the funnel picture still earns its keep as a work plan.

This is the other direction: not your process, but the route people take before they reach it. And mostly, how you find that route out instead of drawing it.

Most maps describe the company, not the customer

The mistake happens early and never gets caught, because the result looks plausible.

A few people sit down and describe how a customer reaches them. Out comes the route the business wishes for: home page, service page, contact form. Tidy, short, done in one meeting.

Real routes look nothing like that. They contain detours, two-week gaps, a second opinion from outside, and at least one step that happens without you entirely.

The difference is not academic. A map built on assumptions produces fixes for problems nobody has.

What belongs on a map worth having

A customer journey map is not a picture of the route. It is a list of the places where people drop out — and of the questions that go unanswered there.

That is a different job entirely. A picture wants to be complete. A list wants to be useful.

Which is why customer journey mapping usually fails on the beautiful map rather than the wrong one. Diagram every conceivable interaction and you get a wall of sticky notes and not one decision.

Three to five exit points you can evidence beat forty you assume.

The touchpoints you do not own

When businesses inventory their touchpoints, they list what belongs to them: website, newsletter, landing pages, ads, social.

The places where decisions actually turn are often somewhere else.

A review somebody reads without you ever knowing. A comparison on a third-party site. A phone call to a peer in the same industry. An AI answer that mentions your name — or does not.

You cannot control those points. You can know about them and make sure what is there is accurate. That is the difference between a map that lists your channels and one that describes the route.

Where the data comes from: three sources already running

A credible customer journey analysis does not need a research budget. It needs three sources that are already running in almost every business — and the willingness to read them against each other.

Search Console knows the question asked before first contact

It is free, and it is the only source showing the exact wording people arrive with.

The interesting queries are not the ones with plenty of clicks. They are the ones with plenty of impressions and few clicks: the intent lands, the answer does not convince. Those are open questions, and finding them costs nothing but a look — the cheapest search engine optimisation there is.

Analytics shows where the route breaks

Which pages get opened, where the visit ends, which form step stops being completed.

The order matters: find the exit first, look for the reason second. Which metrics are worth watching at this size is covered in the piece on Google Analytics 4.

The first conversation shows what no tool can see

Ask at the start of every first call how the person came across you — and write the answer down verbatim, not sorted into categories.

Twenty of those say more about the route than any attribution model. They contain what no software records: referrals, conversations at events, a post somebody read a year ago.

Add the people in your business who talk to customers daily. They know the recurring objections by heart. That is a data source, not a hunch.

The mistake that wastes all three

Reading each source on its own and deriving one action from each.

The insight sits in the contradiction: a question Search Console shows that nothing on the site answers. Or an objection that comes up in every sales call and appears nowhere in the numbers.

The questions that have to be on the map

Every place gets the same test: what is the person asking here — and does anything answer it?

For services, three questions decide it nearly every time.

"Is this meant for a business like mine?"

The question behind every service page. Leave it unanswered and you lose people who would have been a good fit, because they never felt addressed. References at their own scale answer it faster than any description — which is what case studies are for.

"Why should I believe you?"

More gets decided here than on any ad. What works is names, numbers, faces and checkable claims. What does not work is adjectives.

"What happens if I get in touch?"

The most expensive unanswered question there is. If someone cannot tell whether hitting send produces a sales call, a proposal or a price list, they tend not to hit send.

These three are why the website comes out as the most important point in nearly every analysis: it is the only place where you can answer all three at once.

From map to decision: three patterns

A finished map is worth something only once it produces an action. Three patterns turn up often enough that the diagnosis usually hangs on them.

Plenty of visitors, few enquiries. The route to the page works; the page does not. Check first whether it answers the question people arrive with. The usual causes are in the piece on why a website brings in no customers.

Plenty of service-page views, no enquiry. Proof is missing, not argument. That is an evidence problem — and what else a high-converting website depends on is a subject of its own.

Plenty of enquiries, few contracts. Then the weak point is not in marketing at all. It is in the offer, the price or the follow-up — and without a CRM you cannot see it, because measurement stops at the form.

Telling those three apart is the difference between a sensible fix and an expensive rebuild in the wrong place.

One place, then measure

The instinct after a first analysis is to improve everything at once. That is the most reliable way to know nothing afterwards.

Change five things at once, see enquiries rise, and you cannot say which of the five did it. Next time you start from zero again.

Take the single place where most is lost. Change it there. Wait four weeks. Then the next one.

That is slower than it feels — and faster than it actually is.

A map has an expiry date

Search behaviour shifts, channels appear, and AI answers now sit where a list of results used to be two years ago.

A map that is two years old describes a route that no longer exists in that form. Once a year is enough for most businesses, as long as the three data sources keep running in the background.

How that route is shifting right now is covered in the piece on the future of Google search.

Frequently asked questions

What is the difference between a customer journey and a marketing funnel?

The customer journey describes the route from the customer's side; the funnel describes your own work behind it. Both describe the same stretch from opposite ends — the marketing funnel organises your tasks, the journey organises people's questions.

How long does customer journey mapping take?

Drawing the map takes a day. Collecting the data takes two to four weeks, because twenty first conversations do not happen in one week. The order matters more than the speed.

Does a smaller business need software for this?

No. Search Console, Analytics and written-down answers from sales calls are enough for the first map. Software becomes necessary only when you want to trace an enquiry through to the signed contract.

How many touchpoints belong on the map?

As many as you can evidence. A map with five evidenced points is useful; one with thirty assumed points is not.

In closing

The reflex when enquiries dry up is to turn up the pressure: more channels, more budget, more posts.

Knowing the journey almost always points the other way. Fewer places, but the right ones. Once you know where people drop out and what they are asking at that moment, you do not need five new channels — you need one good answer in a known place. That is what ALPENIQ Growth is built around.

Good marketing does not start with the ad. It starts with the question somebody is holding before your name ever comes up.