CALIBRA™
Volume VII

Pricing Architecture.

What to charge, and what happens when you change it.

The problem

Price is the strongest lever in the business and the least governed.

A one per cent improvement in realised price drops almost entirely to the bottom line. Most businesses know this and still set price by annual increment, competitor observation and whatever the sales team will tolerate discussing.

The gap between list price and realised price is where the money actually goes, through discounts, rebates, terms and exceptions that no single person sees in total. Nobody owns the difference, so nobody manages it.

What the volume does

List, realised, mix and packaging, read as one architecture.

The reading covers the whole price structure rather than the headline number: what is listed, what is actually collected, what the customer and product mix does to the average, and how the offer is packaged. Each goes through the four lenses.

Price changes have consequences the finance model rarely carries: what they do to how the brand is understood, what future positions they close off, and whether the business can operationally sustain the new structure. Those are lenses two, three and four, and they are where most pricing exercises come unstuck.

Pricing runs seven instructions. Raise. Lower. Hold. Trial. Bundle. Unbundle. Reframe. Each carries a timeline, a named owner and a check date. The pricing readings sit against a design target of 0.84 drawn from a development cohort of twenty-four engagements between 2019 and 2026. That is a target rather than an observed client result, and it will be replaced by the recorded figure as the register fills.

How a reading works

How a pricing reading works.

You provide what the business already holds: the price list as it actually operates, the discount and rebate records behind it, contribution by product or tier where finance can produce it, and any customer research on price. Nothing is built specially for the exercise. Where contribution has never been measured at that level, the reading begins by extracting it.

Each pricing position is then read four ways. What it contributes after the cost of operating it. How customers actually place the price when they compare. Which future moves the position keeps open. What could go wrong: customer reaction, competitor response, regulation, execution. The four readings carry explicit weights and combine into a single view, and where they disagree, the disagreement is recorded rather than smoothed over.

What arrives is one instruction per position, drawn from seven: Raise, Lower, Hold, Trial, Bundle, Unbundle or Reframe. Each carries a stated confidence level, a named owner inside the business, and the date the outcome will be checked against the call. That check is committed to before the reading starts, not once the result is in.

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

The price that had not moved because nobody wanted the conversation

Raise
Situation
A specialist industrial distributor, roughly $180 million turnover, on a price list adjusted only for cost inflation for seven years.
They asked
How do we protect margin against input costs?
The reading found
The business had been absorbing value it was not charging for. Service levels, technical support and stock availability had all improved and none of it had ever reached the price.
Instruction
Raise on the technical range by a stated margin, staged across two quarters, with the service position made explicit to customers first.
Why it matters
Cost-plus pricing quietly assumes your value has not changed since the last increase. For most businesses that assumption is wrong in your favour.
Use case 2

The all-inclusive offer subsidising its heaviest users

Unbundle
Situation
A managed IT services provider, roughly $70 million revenue, selling a single flat monthly rate.
They asked
Should we raise the monthly rate?
The reading found
The flat rate averaged across customers whose support consumption differed by a factor of nine. A general increase would have driven out the profitable customers and retained the unprofitable ones.
Instruction
Unbundle support tiers from the base rate. Hold the base, price the consumption.
Why it matters
A blunt increase on an averaged price selects against you. The customers who leave first are the ones you were making money on.
Use case 3

The pricing question that was a positioning question

Reframe
Situation
A premium food brand, roughly $55 million turnover, losing share to a cheaper competitor in the major supermarkets.
They asked
How far do we have to drop to compete?
The reading found
The reading declined to issue a price instruction. The brand's recognition reading had fallen far enough that no price point would restore the position, and a drop would have confirmed the customer's new view of the category.
Instruction
Reframe. Resolve the brand position before touching price. Any price instruction now would be premature.
Why it matters
Price is where declining brands go to die faster. The method refused to supply the number because the number was the wrong instrument.
Use case 4

The bundle that everyone was sure about

Trial
Situation
A telecommunications reseller, roughly $120 million revenue, considering a hardware and service bundle.
They asked
Do we bundle?
The reading found
The reading could not separate the effect of bundling from the promotional period that would launch it, and the available comparison data came from a market with different handset economics.
Instruction
Trial in two states for two quarters, with the exit condition and the measure fixed before launch.
Why it matters
A trial with no stated exit condition is a launch with extra steps. Naming the condition in advance is the only part that matters.
Use case 5

The increase the market would have absorbed and the business could not deliver

Hold
Situation
A contract manufacturer, roughly $240 million turnover, with a strong order book and stretched capacity.
They asked
How much can we put prices up?
The reading found
The first three lenses supported a rise. The fourth did not. Lifting price while service levels were already slipping would have converted a capacity problem into a customer problem.
Instruction
Hold price. Resolve capacity, then re-read within two quarters.
Why it matters
The fourth lens overrules the other three more often in pricing than anywhere else. A price you cannot deliver against is not a price, it is a promise you will break.
Use case 6

The premium that outlived its reason

Lower
Situation
A specialist components manufacturer, roughly $150 million turnover, held its flagship line at a premium of about fifteen per cent, set a decade earlier when the product led its category on a feature every serious rival has since matched.
They asked
Whether the premium could still be defended, and what defending it would take.
The reading found
The financial reading defended the price: contribution per unit was the strongest in the range. The customer reading did not. Buyers could no longer say what the premium paid for, and the volume drifting to the nearest rival was costing more than the premium was protecting. The risk reading added a timing point: a forced retreat later would look like weakness, a chosen one now would not.
Instruction
Lower. Close most of the premium gap in two steps over two quarters, keep a modest margin for the features that still register with buyers, with a named owner and the outcome checked at twelve months.
Why it matters
The method is not a machine for putting prices up. A premium is a claim the customer has to keep agreeing with. When they stop agreeing, the discipline is to move first, on your own timing, rather than be moved slowly by the market.
For the academics

Pricing sits between an econometric tradition estimating elasticities from history and a behavioural tradition explaining why those elasticities keep being wrong, and the two have coexisted for thirty years without either having to concede anything. Neither produces a decision that survives contact with a sales director who has discretion. Volume VII adds a fourth-lens deliverability test and a check date, which mostly serves to show how often the elegant answer was undeliverable.

And commercially: it finds the money between list price and the bank.

The check

Every instruction has a date.

Twelve months after a reading, the instruction is compared against what happened and the result is recorded, whether or not it is flattering.

Start a conversation

One sentence is enough.

A pricing reading settles what a price list cannot tell you on its own: which positions should move, which should hold, and which are answering the wrong question. It ends in one instruction per position, each with a confidence level, a named owner and a check date. It starts smaller. You describe the decision in front of you, and Rob reads whether the method fits it before anyone commits to anything. Send one sentence: what is the decision you are trying to make?