Haute Lumière
Commerce · III.11 · MMXXVI · daylight
For the person with a P&L, a pricing committee, and a set of things the firm gives away without ever having decided to. This chapter is not a caution about markets. It is a pricing discipline with two tests in it, and it will find money in your business within the quarter.
Your firm makes three kinds of decision about price, and it has a rigorous process for only one of them.
Priced goods. Full apparatus: elasticity, cost-to-serve, competitive position, margin. Nobody sets a price here by instinct.
Goods you have decided not to price — free tiers, waived fees, unbilled service hours, internal shared services recharged at zero, employee goodwill, the after-sales help that nobody books. These are allocated by a rule. The rule is almost never written down, nobody computes its incidence, and it reliably favours whoever is already inside: the largest accounts, the loudest customers, the teams nearest the function.
Behaviours you have tried to change with a charge — late fees, minimum order values, expedite premiums, penalty clauses. Here the evidence says something specific and commercially expensive, and it is in this chapter.
This workbook does three things: computes the incidence of your unpriced goods, applies a test that tells you which of them can be priced at all, and installs five non-price instruments you already use informally, so you can use them on purpose.
Exercise 1.1 — The unpriced inventory (one week, with your controller)
List every good or service the firm supplies at zero or below cost-to-serve. Include internal ones. For each, five columns:
| Good | Cost to serve | Who consumes it | Consumption concentration | Why it is free |
|---|
The fourth column is where the work is. Compute what share of consumption goes to your top decile of consumers. In almost every case it will be far above their share of revenue, because a free good is rationed by whoever is best placed to ask for it, and that is your largest and most sophisticated counterparties.
You are running, on your own P&L, the arithmetic the chapter runs on a water utility — where the richest quintile takes 40.3 percent of a $74,646,000 subsidy and the poorest 5.6 percent, a ratio of 7.24.
Exercise 1.2 — The appreciative pricing conversation (one session)
Take the leadership team through three questions, in this order, and resist the fourth for a full hour:
That third question is the one that returns most. Reciprocity in a commercial relationship is a real asset with a real regeneration rate, and it is the asset most easily destroyed by a well-intentioned fee.
Exercise 2.1 — The incidence table (one week)
Pick the largest item from Exercise 1.1 and build its table. Rows: your customer or business-unit deciles. Columns: consumption of the free good, revenue, cost-to-serve absorbed, and the implicit transfer.
Then compute the two ratios the chapter computes:
The board line you are building toward: this free service is a transfer of £X a year, and Y percent of it goes to the counterparties with the strongest negotiating position.
Exercise 2.2 — The convergence test on your intangibles (half a day)
Your firm carries several things it says cannot be valued: brand permission, institutional knowledge, the safety culture, the relationship with a regulator. Some of those can be priced. Some genuinely cannot. Stop guessing which.
For each, two independent methods — replacement cost, foregone margin under a modelled loss, insurance premium, a comparable transaction — and one ratio.
This single exercise ends the annual argument about intangibles, and it ends it with a number that neither side chose in advance.
Exercise 2.3 — The fine that is a price (2 hours)
Audit every penalty, late fee and charge you levy to discourage a behaviour. For each, answer: what was holding this behaviour in place before the charge existed?
If the answer is an obligation, a relationship or a professional norm, the charge is a substitution, not an addition. The Haifa study is the evidence: a fine of 10.00 NIS against a monthly fee of 1,400.00 NIS — 0.714 percent — took late collections from roughly 8.0 a week to roughly 20.0, a multiplier of 2.50, and withdrawing the fine did not bring them back down.
Two commercial rules follow, and both are testable in your own data.
Exercise 3.1 — Install the catalogue (one week)
Each of these is something your firm already does by accident. Do it on purpose.
Quantity. Fix the volume, discover the price. Internal carbon budgets, allocated headcount, capped expedite slots auctioned between units. The public benchmark: the sulphur dioxide allowance market cut covered emissions 36.0 percent between 1990 and 2004 while coal generation rose 25.0 percent — an intensity of 0.5120, a fall of 48.8 percent — at roughly $1 billion a year below the command-and-control alternative. The lesson for a firm: a quantity instrument beats a mandate whenever marginal abatement cost varies widely across your sites, and it usually does.
Standards. Fix the outcome, free the route. The lead phase-out took children's blood lead from 15.0 µg/dL to 2.8 µg/dL — 81.3 percent, a ratio of 5.36 — for an estimated $110,000,000,000 to $319,000,000,000 per birth cohort, $27,500 to $79,750 per child over 4,000,000 births. And it used tradable lead credits inside the standard. A standard with a trading layer is the highest-yield regulatory design in the record, and the same structure works inside a group with multiple plants.
Rights with a clearing mechanism. Where something is genuinely not for sale internally — a person's time, a safety veto, a customer commitment — build the exchange rather than the market. Kidney paired donation clears roughly 1,000 transplants a year at a saving of $56,000 per patient-year: dialysis at about $94,000 against a graft at about $38,000, with a first-year excess of $26,000 recovered in 5.6 months and $477,691 of present value at 3 percent over ten graft-years.
Rationing. Fix the per-unit entitlement. And budget for the cost nobody mentions: Cape Town cut consumption 58.3 percent, from 1,200.0 to 500.0 megalitres a day, and carried a revenue shortfall whose midpoint, R1,700,000,000 against 255,500,000,000 litres saved, is R6.654 per cubic metre. If your business restricts consumption of the thing it sells, the restriction has a P&L line. Put it in the plan before you announce the target.
Deliberation. Put the allocation in a room with a rule. Oregon ranked 709 condition-treatment pairs and funded to line 587 — 82.8 percent, with 122 named below the line — and the first list was rejected and rebuilt. Porto Alegre took water connections from 75.0 percent to 98.0 percent and sewerage from 46.0 percent to 85.0 percent, closing 92.0 percent and 72.2 percent of the respective gaps in nine years. Deliberation's cost is a rewrite, and a rewrite is the instrument working.
Exercise 3.2 — The pricing paper (one week)
For the item chosen in Exercise 2.1, write the board paper:
Exercise 3.3 — The audit conversation (one meeting, early)
Where the reform converts a recurring operating subsidy into a durable asset — the chapter's connection facility is the archetype — take it to your auditors before you take it to the board. Converting a recurring operating expense that purchases nothing into a depreciating asset with a revenue stream attached is a conversation about useful economic life, which they have every year.
Exercise 4.1 — Into the pack (one conversation)
One line in the standing monthly pack: the transfer value of the largest free good and its concentration ratio. Anything reviewed monthly persists; anything reviewed by exception does not.
Exercise 4.2 — The standing question (one cycle)
Add to the pricing committee's template: if we are choosing not to charge for this, who receives the benefit and who pays the implicit charge? It costs one line and it will stop at least one expensive decision a year.
Exercise 4.3 — The non-convertible recognition (one cycle)
Wherever you were about to attach a small payment to a discretionary behaviour — referrals, knowledge sharing, mentoring, safety reporting — test a recognition that cannot be converted to cash first. The Swedish result is the evidence: women's donation fell from 52.0 percent to 30.0 percent under a small payment, a 42.3 percent relative fall, and the effect vanished entirely when the money could be assigned to charity. Run it as a split test across two regions. The cost of the test is a quarter.
Exercise 4.4 — The one page, day 90
Transfer value, concentration ratio, convergence result, instrument chosen, cost named, decision number. One page, one person, the one who controls the pricing mandate.
Exercise 4.5 — Delight, for a firm (ongoing)
The pleasure available here is specific: a pricing committee that stops having the same argument. The convergence test settles in twenty minutes a question that has been recurring for years, and it settles it without anyone having been wrong about their values. Watch for the meeting where somebody who resisted this uses the concentration ratio in their own argument. That is the moment it has stopped being your initiative.
The incidence table built on data you do not hold. Very common. The honest entry is unknown, never flat. A fallback that turns "we could not find out" into "the answer is nothing" is how a pricing committee starts lying to itself.
The convergence test run by someone who has decided. Two methods can be chosen to agree. Have the test run by a function with no stake in the answer, and publish both methods.
The crowding-out finding carried out of its domain. The Swedish result is about small sums, a visible conversion of a gift into a sale, and one demographic. It does not travel to a gift card at a blood drive, where incentives raised turnout. Do not build a company-wide policy on one study.
The word "incommensurable" as a door. Once the vocabulary arrives without the arithmetic, this cannot be priced becomes the most convenient sentence in the building. The test is the antidote, and the test must be mandatory.
Cost-reflective pricing without a floor block. This is the one that ends careers. The reform is progressive only when the rebate or the lifeline block is on the first page of the model.
| Day | Action | Artifact |
|---|---|---|
| 1–15 | Unpriced inventory with cost-to-serve | The list, costed |
| 16–30 | Appreciative pricing conversation | Three answers, written |
| 31–45 | Incidence table and concentration ratio | The table |
| 46–55 | Convergence test on the intangibles | Two methods, one ratio each |
| 56–70 | Instrument chosen; floor block designed | The pricing paper |
| 71–80 | Audit conversation on useful economic life | Treatment agreed |
| 81–90 | Into the pack; the one page | The decision number |
Subject. Pricing and allocation of [good], currently supplied at zero.
The transfer. [good] costs £X a year to serve and is supplied without charge. The top decile of consumers takes Y percent of it against Z percent of revenue — a concentration ratio of R.
The test. Two independent valuation methods return [A] and [B], a ratio of [A/B]. The good is therefore [priceable / not priceable].
The proposal. [Tariff with a floor block of N units at zero] or [instrument, from the catalogue], at a stated cost of £C.
The incidence of the proposal. Decile 1 net [+/−], decile 10 net [+/−]. The reform is [progressive / regressive] on the first page of the model.
The number that decides it. [One inequality.]
What we are not claiming. [Where the data is unknown rather than flat.]