The methodology

The Barrier Matrix.

The Barrier Matrix is a go-to-market methodology. It maps a B2B SaaS customer journey to six phases, then checks every phase against the three reasons people leave: confusion, then fear, then friction, always in that order. Eighteen cells. Each one either clean, or named with a severity and the behavior behind it.

This page is the whole method, given away. If you want to run it yourself, everything you need is below.

Why it exists

Your analytics tells you where. It cannot tell you why.

A funnel chart records three people leaving the same page and calls them the same event. One didn’t understand what you sell. One understood it and didn’t trust it. One wanted it and hit something in the way. Three reasons, three different fixes, one identical number.

So the fix becomes a guess, and the guess is usually a redesign, and when the number doesn’t move the conclusion is that conversion work doesn’t work. What actually happened is that a confused buyer was handed reassurance, or a frightened one was handed a shorter form.

The matrix is a structure that forces the question which of the three is this before anyone is allowed to propose a fix. That is the entire idea. It turned a hunch into a checklist you can argue with.

The first axis

Six phases, and most companies measure two.

A funnel starts when someone arrives. A customer journey starts earlier — when something made their current setup feel wrong, and through the months they spent half-looking without ever searching for you. Two of these six phases happen before you can measure anything, and one happens after the money has already been paid.

01

First thought

Something makes the current setup feel wrong.

Nobody is searching yet. A number slips, a process breaks, someone new arrives with a different standard. The decision starts here, months before any of it reaches your analytics.

Fails as: You are invisible at the moment the problem becomes real, so the shortlist forms without you on it.

02

Passive looking

They are curious, not comparing.

Half-attention. A newsletter, a post someone shared, a conversation at a conference. They are collecting a sense of what exists, without any intention to buy this quarter.

Fails as: Nothing of yours is where they browse, so when the search does start, you are not one of the names they already half-know.

03

Active looking

They are shortlisting, and you are on it.

Now they compare. Your site, two competitors, a review site, whatever a colleague recommended. This is the first phase most companies measure, and the third of six.

Fails as: They cannot work out what you actually do or who you are for, so they drop you rather than risk being wrong about you.

04

Deciding

They want it and have to justify it.

The pricing page, the demo, the internal conversation you never see. Someone has to defend this choice to a colleague or a budget holder, using only the material you gave them.

Fails as: The downside is unmanaged. What happens if this does not work is unanswered, so the safest move is to do nothing.

05

First use

They have paid and nothing has happened yet.

Setup, integrations, getting one real outcome. The gap between signing and the first moment the thing is useful is where most of the money that has already been spent quietly goes.

Fails as: They stall halfway through setup and never reach the point the purchase was for.

06

Ongoing use

It works, and they stop noticing.

Renewal, expansion, or a quiet exit. Value delivered but never surfaced is value that does not get remembered when the invoice arrives.

Fails as: They cannot see what they got, so the renewal becomes a fresh decision rather than a continuation.

The second axis

Three reasons people leave. Only three.

Every drop-out that isn’t a traffic or product problem is one of these. The claim that there are exactly three is the useful part: it is small enough to check every phase against, and specific enough that the three demand different fixes.

01 · Confusion

“I don’t understand what this is.”

Cognitive load, ambiguity aversion, choice overload.

They cannot form a clear picture of what they would get, so they postpone. Postponing is not a decision, which is why it never looks like one in your reporting — it looks like disinterest, and disinterest gets blamed on traffic quality.

The test: Could a stranger describe your offer back to you in one sentence after thirty seconds on the page?

02 · Fear

“I understand it. I don’t trust it.”

Loss aversion, social proof, the endowment effect.

They have understood the offer and are weighing what it costs them to be wrong — professionally, not just financially. No amount of extra clarity fixes an unmanaged risk. Explaining harder is the standard response and it does nothing.

The test: Is the exit visible before the commitment? Can they see what happens if this goes badly?

03 · Friction

“I want it. Something is in the way.”

Steps, fields, waiting, and not knowing what happens next.

They already want to act. This is the cheapest of the three to fix and the one nearly everyone starts with, because it is the only one you can see without asking anybody anything.

The test: Count the steps between wanting it and having it. Then count the ones that exist for your convenience.

The order is the method

Confusion, then fear, then friction. Never together.

If you take one thing from this page, take the order. It is the part that makes the difference, and it is the part everyone gets wrong, because the three announce themselves in exactly the reverse of their importance.

Reassurance lands on a confused buyer as noise. Testimonials, guarantees and security badges all answer a question someone who doesn’t understand your offer hasn’t reached yet. You are managing a risk they cannot evaluate.

A shorter form doesn’t help someone afraid of the decision. Removing three fields from a signup is real work with a measurable output, which is why it is so often the first thing done, and why it so often changes nothing.

Run them at once and you get a list of everything wrong with your site, sorted by how easy each item is to fix. Run them in order, one pass a day, and you get a diagnosis. Doing all three at the same time is how the wrong reason gets the blame.

There is one more reason for the sequence. Fixing confusion changes what the other two look like. Once people understand the offer, different objections surface and different steps start to matter. Diagnosing friction on top of unresolved confusion means diagnosing a journey that will not exist after the first fix ships.

Eighteen cells

Six phases across, three reasons down.

Six times three is where the eighteen comes from. Every cell gets asked the same question: at this phase, for this reason, is anything going wrong? A cell is either clean or it holds a finding with a severity and the behavior driving it.

Read down a column and you see which phase is losing people. Read across a row and you see which reason you have been misdiagnosing. Empty cells matter as much as full ones — they are where you can stop spending attention.

First Thought
Passive Looking
Active Looking
Deciding
First Use
Ongoing Use
Confusion
1
2
4
6
2
Fear
1
2
5
3
1
Friction
1
3
4
2
Swipe to see all six phases

An illustrative matrix. Yours is built from your journey, your analytics and your customers’ words.

Drop-outsCritical

The six phases, in order: first thought, passive looking, active looking, deciding, first use, ongoing use.

Run it yourself

Seven steps, and nothing here is proprietary.

This is what I do, in the order I do it. The method is not the moat; the discipline of following it when you already have a favorite theory is. If you run it yourself and it works, that is a good outcome.

  1. 1

    Write down what you already think is wrong.

    Before you look at anything. Date it. This is the part that makes the rest honest — without a baseline, every finding feels like something you already knew.

  2. 2

    Collect evidence and conclude nothing.

    Analytics exports, call recordings, support tickets, cancellation reasons, every page someone passes through. The discipline is separating collection from interpretation; doing both at once means you find what you expected.

  3. 3

    Make one pass for confusion. Only confusion.

    Walk all six phases asking a single question: could someone fail to understand what this is here? Note every instance. Do not fix anything yet.

  4. 4

    Make a second pass for fear.

    Same six phases, new question: what is being risked here, and is that risk answered? Fear findings cluster in deciding, and they are the ones that look like pricing problems.

  5. 5

    Make a third pass for friction.

    Last. Steps, fields, waits, dead ends. It comes last because friction is the loudest and would otherwise drown out the other two.

  6. 6

    Fill in the eighteen cells and rate severity.

    Each cell is clean or named, with a severity and the behavior behind it. Then read the grid: the heaviest column is the phase losing you the most, and the heaviest row is the reason you have been misdiagnosing.

  7. 7

    Fix in order. Confusion first.

    Not the easiest first, and not the loudest. Out of order, the work fails and you conclude the finding was wrong when the sequence was wrong.

What it doesn’t do

Where the method stops.

A methodology that explains everything explains nothing. Four things this one genuinely cannot do.

It cannot fix a traffic problem

If almost nobody arrives, there is nothing to diagnose. Eighteen cells over a hundred visitors a month is numerology. Below roughly €5k a month in acquisition spend, the maths does not work.

It cannot fix a product problem

If people understand it, trust it, reach it and then leave because it does not do the job, that is not a journey problem. The method will tell you that, which is useful, but it is not what it is for.

It is judgement, not a score

Severity is assigned by a person. Two practitioners will differ at the margins. What the structure removes is not judgement but the freedom to skip a question you would rather not ask.

It is a diagnosis, not a test

It tells you why, with evidence, before you build. It does not prove causation the way a properly powered experiment does. Most companies do not have the traffic to run those anyway, which is exactly why guessing has been the alternative.

Where it came from

I built it because I kept fixing the wrong thing.

For years my go-to-market work looked like this. Find the biggest drop-out. Form a theory about it. Ship a change. Watch the number not move. The theory was always plausible — that was the problem. Nothing told me which of five plausible explanations to test first, so I tested the one I happened to think of.

The phases come from how buying decisions actually run, which is nothing like a funnel. The three reasons come from behavioral science that has been sitting in the literature for decades — loss aversion, ambiguity aversion, cognitive load — none of which I invented and most of which never gets applied to a journey in any systematic way. The contribution here is the grid and the order, not the psychology.

More on the person behind it on the about page, and the engagement that applies it, day by day, on the services page.

Or have me run it on your journey.

Seven working days, under an hour of your time, and the matrix comes back filled in for your customers rather than an illustration.

See the price