One customer, four versions

Do not index

The Predictive Path

Course 1: Strategy without control

Lesson 4: One customer, four versions

One customer, four versions

Why the same account is four different things, seen at four different points, and never assembled into one.

Inside your company, one customer is four different customers: a lead to marketing, a deal to sales, an account to customer success, recognised revenue to finance. Four definitions, four starting points, four moments — and they never line up. This lesson is about what falls into the gaps between them.

One customer, four answers

Ask marketing about a customer and you hear that it was a strong logo from a high-intent segment. A good acquisition. Ask sales and you hear it closed at a healthy contract value with a clean structure. A good deal. Ask customer success and you hear adoption is steady and the relationship is calm. A good account. Ask finance and you hear the margin is below the segment average and the cash arrives slowly. A marginal customer.

None of them is wrong. They are describing the same customer, and the readings do not agree.

They do not agree because they were never taken at the same moment. Marketing's read is from acquisition. Sales's is from signature. Customer success's is from somewhere in the middle of the relationship. Finance's is from much later, once the economics resolve. Four true readings of one customer, taken at four different points, and nobody is looking at the same customer the others saw.

Most revenue organizations carry one of those readings — usually where the customer sits in the process right now — and treat it as the customer. It is not. It is one reading, taken at one point, of something that is four different things at four different times.

The customer moves along a lifecycle

The reason the readings separate is that a customer is not a position in a process. It is something that moves along a lifecycle.

The lifecycle is the path the customer actually travels: it is acquired from some segment at some cost, it signs a contract with some structure, it adopts the product over months or it does not, it expands or contracts, it costs something to serve that changes over time, it renews or it churns. That path is the same path for every customer. What differs is what happens at each point along it.

Four facts about the customer get decided at four different points on that path:

The demand fact — which segment it came from, what it cost to acquire, what intent it carried. Settled at the start, before the deal.

The deal fact — the contract structure, term, pricing, discount, support tier. Fixed at one point, signature, and then governing everything that follows for years.

The usage fact — whether the product is adopted, how deeply, by how many, how durably. Resolved slowly, across the long middle of the lifecycle.

The margin fact — what the relationship costs to serve against what it pays, and how that gap moves. Resolved last, well after every decision that set it.

These are not four views of one moment. They are four facts that become true at four different points along the path. At the point marketing reads the customer, only the demand fact exists. At signature, the deal fact is set but the usage and margin facts have not happened yet. By the time the margin fact is legible, the demand and deal facts are long behind and the decisions that drove them are forgotten.

The funnel is good at one stretch of the path

The funnel is the model most organizations still run on, and it earned its place honestly.

A funnel is an excellent instrument for one stretch of the lifecycle: the part where a prospect moves through commitment toward a signed deal. Stages, conversion, velocity, where deals fall out. For the deal stretch of the path, a funnel is the right tool and works well. The same is true elsewhere in pieces — a renewal pipeline is a funnel, an expansion pipeline is a funnel, and each is useful for its own stretch.

The problem is not the funnel. It is the assumption that a funnel built for one stretch describes the whole path. It does not, because the facts that decide whether the customer was worth having — margin, cash, expansion, retention — do not resolve in the deal stretch. They resolve at other points along the lifecycle, points no single funnel spans. A sales funnel cannot hold margin, because margin is not decided where the funnel is looking. It is decided later, somewhere the funnel does not reach.

The cost is set early and read late

Put rough numbers on one customer and follow the path, not for precision but for the shape.

At the start, the demand fact is complete: the customer came from a segment, at an acquisition cost, with an intent profile. At signature, the deal fact is set — a term, a structure, a support tier — and then it does not move for twenty-four or thirty-six months while the rest of the path runs. Across the long middle, the usage fact unfolds: adoption builds or it does not, deepens across teams or stays with one. Underneath all of it, resolving last, the margin fact: cost-to-serve against what the customer pays, a gap that does not read clearly until the relationship has roughly eighteen months behind it.

So at six months into the path, "how is this customer doing?" has four honest answers from four points: a demand answer that is already history, a deal answer frozen at signature, a usage answer that is real but not yet conclusive, and a margin answer that barely exists yet. Whatever instrument the organization is using reports the point it is standing on. The fact that decides whether this was a good customer — margin — is the one that will not be legible until long after every decision that set it was made.

The economics are not lost in a moment of failure. They are set early, at points the organization was not reading, and discovered late, at the one point it was. In between, every reading available looked fine, because the readings were measuring the points that were fine and were silent, by construction, about the points that were not.

Nobody can assemble the whole customer

There is a sharper problem underneath the four facts, and it is worth stating plainly.

The four readings are not just taken at different points. They are taken by different functions, into different tools, using different definitions of the same words. Marketing's "good customer" is a segment-and-intent definition. Sales's is a contract-value definition. Customer success's is an adoption definition. Finance's is a margin definition. The data behind each one sits where that function put it, shaped the way that function needed it. It is not connected across the points. It does not accumulate into one record of the customer as it moves.

So there is no point on the path where the whole customer can be seen. Not because no one has looked, but because the readings never join. Each function sees its slice, at its point, in its own terms, and the slices are never assembled into one moving picture of the customer the path actually produced.

And because the slices never join, nothing learns. When a customer churns, or turns out to carry thin margin, the organization does not read the whole path back to see what the demand fact or the deal fact had already determined about this outcome. It cannot — the path was never recorded as one thing. So the same segment gets acquired again, on the same signals, and the same outcome forms again, because nothing fed the end of the path back to the start of it. What a revenue built to read its own lifecycle back into its decisions looks like is exactly where the next course begins.

It has to be right at the lowest level or it is not right anywhere

Leadership rarely asks about one customer's path. It asks about the segment and the company. How is the enterprise segment trending. What does the renewal base look like next year. Where will the company land against plan. Those are the questions that get reported upward and acted on.

None of those higher-level readings is a separate measurement. The segment path is the customer paths in that segment, added up. The company path is every customer path, added up. There is no segment lifecycle or company forecast that exists independently of the customer-level readings underneath it — they are those readings, aggregated.

Which means the error does not stay where it started. If the customer-level reading is wrong — the demand fact mislabeled, the margin fact not yet legible, the slices never joined — the segment path built on it is wrong by the same amount, and the company path built on that is wrong again. Aggregation does not average the error out. It carries it up. A company forecast assembled from customer readings that do not hold is not a steady number with some noise in it. It is an unsteady number that looks steady because the instability is one level down, out of view.

This is why the foundation has to be built from the ground up. Not because the lowest level is the most important reading in itself, but because every reading above it inherits whatever is true of it. Get the customer level wrong and there is no level at which the company becomes clear again — the segment and company paths are unsteerable for the same reason the customer one was, only harder to see. What it takes to make the lowest level sound enough to build on is taken up in the next course.

The reframe Course 1 has been building toward

This is where the first three lessons converge.

Lesson 1.1: seeing is not steering. Lesson 1.2: a result you cannot explain is not one you can repeat. Lesson 1.3: each function answers for consequences it cannot see forming. The same fault sits under all three, and now it has a precise shape.

The customer moves along a lifecycle. The facts that decide its worth are set at different points along that path, read by different functions, in different terms, and never assembled. A customer acquired in a strong demand quarter, signed on a structure that compresses lifetime margin, adopting steadily but expensively, reads as a clean win at every point the organization actually looks — because no point it looks at can see the path whole, and nothing carries one point's reading to the next.

What this lesson does and does not claim

Be exact about the boundary, because the next move is the tempting one and it belongs to the next course.

This lesson names a problem. A customer is four facts decided at four points along a lifecycle, read by four functions that never assemble them, and a single picture taken at one point cannot hold the path. That is the limit of what Course 1 sets out to do — name the problem precisely.

It does not describe what a revenue would have to be for the lifecycle to be read as one thing, for the picture to widen at each point instead of resetting, for the end of the path to inform the start of it. That is a real question with a real answer, and the answer is the work the next course exists to do. Reaching for it here would skip the step the rest of the Academy is built to take properly.

The funnel does not disappear in that answer. It keeps doing the one thing it does well — measuring motion through the deal stretch of the path. What changes is that it stops being mistaken for the whole, and takes its place as one good instrument for one stretch of a lifecycle that runs much further than it can see.

Next up

If a customer is four facts along a lifecycle and no point on the path can see them whole, the missing piece is not a better vantage point. It is treating the lifecycle itself as the thing to be reasoned about. That is where the next course begins.

→ Continue to The strategy that stays on the slide

———

This article is part of The Predictive Path

By Niko Laine, SaaS CFO

→ About the author

→ LinkedIn

You can read it on its own or explore the full curriculum.

→ View The Predictive Path

→ Explore courses