CALIBRA™
Volume VI

Distribution and Channel.

Which routes to market to back, which to leave, and how you will know whether the answer was right.

The problem

Channel weights are inherited, and defended by whoever owns them.

Most businesses reach their market through several routes at once, and the weight across those routes was set years ago by decisions nobody now remembers. Each channel has an internal owner who can make a strong case for it. None of them is asked to make the case against.

The result is a channel architecture that accumulated rather than one that was chosen, with cross-subsidies inside it that nobody has looked for.

What the volume does

Every route read against every other route.

Each channel position is read through the same four lenses: does it pay, does it build recognition, does it open future options, and can this organisation actually run it. The point is that they are read together. A channel that looks healthy alone often looks different beside the one funding it.

The seven channel instructions. Expand. Reallocate. Hold. Trial. Exit. Insource. Reframe. Reallocate is the most common, issued in roughly 31 readings in a hundred. Exit and Insource are rare at about four each. Reframe, the instruction to go back and question the framing, is rarest at around two.

How a reading works

How a channel reading works.

You provide what you already have: revenue and cost by route to market, the contracts sitting behind each one, and whatever customer data exists. Nothing new gets commissioned. If the numbers are more than six months old, the reading says so and leans on them less.

Each channel is then read through four lenses. What it earns once every cost is loaded onto it, not just the invoice margin. Whether customers actually look for you there. What it makes possible later. What could take it away. The four scores combine into a single view per channel, carried with a range that states plainly how sure the reading is. A wide range is an answer in itself: it usually means trial before committing.

What arrives is one instruction per channel, drawn from seven: expand, reallocate, hold, trial, exit, insource or reframe. Each comes with a confidence level, a named owner inside your business, and the date the call will be checked against what actually happened.

Six worked situations

What a reading actually produces.

Illustrative composites, drawn from the pattern of engagements rather than from any identifiable client. Figures are indicative.

Use case 1

Three routes to market, one of them quietly subsidising the others

Reallocate
Situation
A food manufacturer, roughly $90 million turnover, selling through supermarkets, food service and a small direct channel.
They asked
Should we push harder into direct-to-consumer?
The reading found
Direct was the most profitable per unit and the least scalable. Food service was carrying fixed costs that supermarket volume was paying for. Read separately each channel looked fine. Read together the cross-subsidy was obvious.
Instruction
Reallocate effort from direct to food service. Hold supermarket. Direct stays, capped.
Why it matters
The most common instruction in channel work, issued in roughly a third of readings. Usually the answer is not a new channel, it is the weight across the existing ones.
Use case 2

The distributor who had become the customer

Insource
Situation
An industrial equipment business selling through a national distributor for nineteen years.
They asked
How do we get the distributor to sell more?
The reading found
The distributor owned the customer relationships, the service data and the renewal cycle. The manufacturer had become a supplier to its own market and could not see its own end users.
Instruction
Insource the service and renewal relationship over eighteen months. Leave initial sale with the distributor.
Why it matters
Insource is issued in about four readings in a hundred, because it is expensive and slow. It is also occasionally the only answer that survives the fourth lens.
Use case 3

The marketplace that looked like growth

Exit
Situation
A consumer brand with rising volume through a global online marketplace.
They asked
How do we grow marketplace sales further?
The reading found
Marketplace volume was growing and margin after fees, returns and advertising was slightly negative. It was also training customers to buy the category on price and hiding them behind the platform.
Instruction
Exit the marketplace over two quarters. Redirect to owned and specialist retail.
Why it matters
Volume growth and value growth are different things. A channel reading that only looks at revenue will get this wrong every time.
Use case 4

The channel question that was not a channel question

Reframe
Situation
A professional services firm debating whether to open offices in two additional cities.
They asked
Which city first?
The reading found
The reading stopped before issuing an instruction. The firm's customers did not choose by geography, they chose by referral. The question assumed a distribution model the market did not use.
Instruction
Reframe. Re-read the customer-side architecture before any channel commitment.
Why it matters
The rarest instruction, about two in a hundred. It is the method declining to answer a well-formed question because the question is wrong, which is more valuable than a confident answer to the wrong thing.
Use case 5

The channel everyone had an opinion about and nobody had tested

Trial
Situation
A household goods business considering a partnership with a large retail group, championed internally and resisted by the sales team.
They asked
Do we sign the partnership?
The reading found
Both camps had a case and neither had evidence. The reading found genuine uncertainty rather than a hidden right answer.
Instruction
Trial with named exit conditions, a fixed window and an agreed measure before it starts.
Why it matters
The instruction is a test, not a decision. Naming the exit conditions in advance is what stops a trial quietly becoming permanent.
Use case 6

The channel being wound down for its age, not its numbers

Expand
Situation
A maker of agricultural machinery parts, roughly $150 million turnover, sold through about eighty independent rural dealers and a three-year-old direct online store. The capital plan and the board's attention had both moved online. Dealer appointments had stopped six years earlier and the network was thinning by attrition.
They asked
Whether to formalise the dealer wind-down and shift the freed funds to the online store.
The reading found
The reading put the dealer network at the top of the portfolio, and not narrowly. It earned the most once every cost was loaded on, dealer customers stayed longer and bought across more of the range, and it was the only channel where the reading was confident rather than hopeful. The online store read mid-range with a wide spread: promising, unproven. The planned wind-down had the two channels the wrong way round.
Instruction
Expand the dealer network. Appoint into the three uncovered regions over twelve months, hold online funding where it is, named owner, checked at twelve months.
Why it matters
The newest channel gets the attention and the oldest one gets the assumptions. A reading weighs both on the same scales, and age is not one of the measures.
For the academics

Channel research is overwhelmingly case-based, retrospective, and written by people with no stake in what happens next. Volume VI reads the entire channel architecture prospectively and at once, which is the only way cross-subsidy becomes visible: you cannot detect a subsidy by examining one side of it, and per-channel analysis is structurally incapable of looking at both.

And commercially: it finds the channel quietly paying for the others.

The check

Every instruction has a date.

Twelve months after a reading, the instruction is compared against what happened and the result is recorded. For channel work the method is built to a design target of 0.83, drawn from a development cohort of twenty-eight readings. That is a target, not a result, and it will be replaced by the observed figure once the register is running.

Start a conversation

One sentence is enough.

A channel reading settles what each of your routes to market actually earns once every cost lands where it belongs, and what to do about each one: a single instruction, a stated confidence, a named owner and a check date. It starts with a conversation, not a document request. You describe the decision in front of you, and Rob reads whether the method fits. If it does not, he will say so. Send one sentence: what is the decision you are trying to make?