Ask a marketing team for their B2B buyer journey map and you usually get a slide. A row of boxes with arrows between them, a verb under each one. Awareness. Consideration. Decision. Somebody drew how buying ought to work. Odds are nobody in that room built it by listening to a buyer. Most people know that, which is why the slide gets shown once and then quietly ignored.
You already own better material. It is sitting in your call recordings.
The map on the slide came from a template
Trace where those stages usually come from. Someone opened a vendor’s diagram and changed the labels to match the product, which kept the shape and skipped the evidence entirely.
You can tell, because the map never says anything surprising. A real journey has a moment where the security team arrives and everything stops while they review. It has a competitor the buyer had already used at a previous job. It has one question nobody on your side thinks is important. None of that survives a template.
Most of the shortlist is settled before your first call
Data table
| Finding | Share of buyers or buying groups |
|---|---|
| Knew a vendor on the list | 97% |
| Winner on Day One shortlist | 95% |
| Shortlist ranked before contact | 94% |
| Knew the winning vendor | 85% |
| Buyer made first contact | 79% |
| Winner spoken to first | 77% |
6sense went and asked the buyers. Their answers are the reason the call matters so much.
6sense surveyed close to 4,000 B2B buyers for its 2025 Buyer Experience Report. North America, APAC and EMEA, median purchase between $200,000 and $300,000. Read in August 2026, it describes a shortlist that is largely settled before any vendor is in the conversation. Buyers make first contact at about 61 percent of the way through. The year before, that figure was 69 percent.
Count what is already true by that point. In 94 percent of cases the buying group had ranked its shortlist. The eventual winner was sitting on that Day One shortlist 95 percent of the time. These vendors were rarely strangers: 97 percent of buyers already had experience with someone on their list, and for 85 percent that included the one that went on to win. Most were not waiting to be found either, because buyers opened the conversation themselves 79 percent of the time. Separately, 77 percent spoke to the winner before they spoke to anyone else.
Read those together and the job changes. You are not guiding a journey. You are joining one late, and asking what happened.
The call is a recording of a room you were not in
This is the part teams miss. A discovery call is not only a qualification exercise. It is the one artefact you own where a buyer describes months of activity you never saw.
They tell you who raised the problem first. Somewhere in there is the tool they tried in March, and the person who has to sign. You hear which competitor they already spoke to, and often why that went nowhere. Analytics gives you none of it. None of it happened on your website.
Read the recording, not the CRM note
Start with transcripts, not summaries. A CRM note is a rep’s compression of a call. Written for the pipeline review, an hour later, from memory. It keeps the budget, the timeline and the next step. It throws away the sentence you actually want, which is usually a passing remark about how the whole thing started.
We start at twenty to thirty calls. That is our house rule and not a validated sample size, so treat it as a starting point rather than a threshold. Spread them across won, lost and stalled. A map built from wins only will tell you that everything you did worked.
Mark five things in every transcript
Code each call the same way, so the reading stays comparable. This is the whole apparatus. A shared document will do.
| What to mark | What it sounds like on the call | What it tells you |
|---|---|---|
| The trigger | “We only started looking after the audit.” | The event that opened the buying window. Compare it with the one your campaigns are built around. |
| The first move | “I asked two people in a Slack group.” | Where the search started. Write it down even when it was not a search engine. |
| The blocker | “Legal sat on it for a month.” | A month lost. Which asset would have bought it back? |
| The unexpected person | “Our data protection officer wants a word.” | Somebody in the buying group your CRM has never met |
| The question you cannot answer | “How does this behave with our ERP?” | The gap between what you publish and what they needed to know |
Mark the words, not your reading of them. Quotes stay quotes.
Let the questions name the stages
Now do the part the template skipped. Put every marked question in one list. Sort it by what the person needed to know, ignoring which call it came from.
Clusters appear on their own. One group asks whether the problem is even worth solving. Another asks whether a category of tool can fix it. A third asks whether your specific product survives contact with their stack. The last one is about defending the purchase to somebody who sat in none of the meetings. Those clusters are your stages. Name them in the buyer’s words. If two of your five drawn stages collapse into one cluster, they were one stage all along.
Keep the vocabulary too. The phrases buyers reach for on calls are often not the ones your product marketing uses. That gap is a keyword research input as much as a messaging one.
What changes on Monday
A map built this way earns its place by moving specific work. Not by looking better on a slide.
The blockers column tells you what to write next. It usually names something dull and load bearing, like an implementation timeline, or a page a security reviewer can read without asking you for it. The unexpected people show you who your demand generation programme is currently ignoring. The unanswerable questions become an FAQ a rep can send instead of booking another call. And the triggers tell you when to be visible at all. That is the difference between a campaign that lands inside a buying window and one that lands in a quiet quarter.
Where this method is weak
Three limits, stated plainly. A method that hides them produces confident nonsense.
You only recorded the buyers who reached a call. Everyone who ruled you out earlier is invisible here. People also narrate their own decisions more tidily than they made them, so treat the story as a reconstruction rather than a log. And a batch this size is a set of anecdotes with a spreadsheet around it. Far better than the template. Still not a statistic. Say “in the calls we read” rather than “buyers do”.
We re-read a fresh batch each time the market feels different, and in practice that lands somewhere around twice a year. Pick your own interval and keep it.
None of that makes the method optional. It makes it honest about what it is: the best record you have of a process you did not witness.
FAQ: the B2B buyer journey
What is a B2B buyer journey?
It is the sequence of things a buying group does between noticing a problem and signing for a solution. Several people are usually involved, and it starts well before anyone fills in a form.
How many calls do I need before the map is usable?
We start at twenty to thirty, which is a house rule rather than a validated number. Mix won, lost and stalled deals. Otherwise the map only describes the path that already worked.
Can I do this without call recording software?
Yes, though it costs more time. Sit in on live calls with a notebook and mark the same five things. The method is the coding scheme, not the tooling.
Do buyers really decide before they contact a vendor?
The shortlist usually is. 6sense puts first contact at around 61 percent of the journey, and reports that 94 percent of buying groups had ranked their shortlist before engaging any vendor.
Does that mean sales calls come too late to matter?
No. Buyers spoke to the eventual winner first 77 percent of the time, which tells you about contact order rather than about what the call achieved. And where the shortlist had not been pre-ranked, the first vendor won only 57 percent of the time.
How is this different from a customer journey map?
Scope. A customer journey map usually covers the experience after purchase. This one stops at signature, and it is built for a group rather than one person.
Who should read the transcripts?
Someone from marketing, not only the reps who ran the calls. We get more out of it when the reader has no stake in how the deal was handled.
How often should the map be rebuilt?
Refresh it with a new batch of calls whenever the market feels different. We land on roughly twice a year. That is our own cadence, not a researched interval.
COSEOM Team
COSEOM is a multilingual B2B growth agency. Since 2008 the COSEOM team has run international SEO, generative engine optimization, paid media and demand generation programs for B2B software and technology companies, in English, German, Spanish and Portuguese. This account publishes our research, playbooks and product guides. More about COSEOM