Pricing

What question does this solve?

Once it becomes viable to sell a discovery, a prototype or a single feature, the provider has to decide what to charge for and how. Pricing by the hour hides whether a delivery exists; pricing by result can hide which result is meant. The model needs to keep price, cost, estimate and value apart, and to say what exactly a price pays for.

Definition and role

Pricing establishes price and charging in relation to the contracted object, the commitment, and the allocation of responsibilities. It expresses in the service what the underlying hypothesis makes possible, without prescribing a single commercial arrangement.

The hypothesis comes before price: if AI reduces the cost of understanding, experimenting, offering, producing and evaluating, new portions of delivery could become economically contractable. Pricing is how that possibility appears in a relationship.

Price, cost and value

Concept	What it is	Relation to the others
Cost	Resources consumed to operate and deliver	Lower production cost does not automatically lower the price
Price	The agreed commercial consideration	It can be assumed with residual uncertainty about cost
Value	The usefulness of the result in the buyer’s context	High perceived value does not show the service can deliver sustainably
Estimate	The provider’s forecast	It informs the price; it is not the price. See Progressive Estimation

How any gains are shared depends on the agreement and on the conditions of supply and demand. The model proposes no market prices, ideal margins or universal formula.

How it works

What is being charged for

Before setting a price, it must be clear what is being bought: available capacity, consumption of resources, a bounded delivery, or a combination. A Delivery Unit is not necessarily a unit of charge. One agreement can cover several units, and one unit can have distinct commercial commitments along its route. A Delivery Stage need not be an instalment either; splitting work operationally can help execution and evaluation without creating independent commercial objects. Calling a price “outcome-based” does not say whether it pays for accepted software, service availability or an observed business benefit.

Possible arrangements

These are design possibilities, not an exhaustive classification or a recommendation.

Possible arrangement	Main object	What must be made clear
Capacity or effort	Availability or work of resources	What is included, and how consumption relates to delivery.
Bounded delivery	A result with defined scope and acceptance	How changes, dependencies and correcting divergences are handled.
Consumption with a limit	Resources used within authorized conditions	The limit does not, by itself, guarantee the delivery will be completed.
Subscription or reservation	Recurring access to specified capacity or service	Limits, priority, availability and rules of use.
Variable share tied to benefit	An observable effect linked to using the solution	Metric, period, attribution, and factors outside the provider’s control.
Hybrid	An explicit combination of objects and responsibilities	How to avoid overlap or ambiguity between the components.

The model does not require a subscription, a fixed catalogue or a standard price per unit.

Pricing a bounded delivery

A defined price for a delivery needs a reference of scope and acceptance, or the parties will read differently what it includes. It does not remove uncertainty; the agreement must say how uncertainty will be handled and which responsibilities stay with each party. Correcting a divergence from a criterion in force is not automatically an extra request, and changing the condition just to turn a failure into a new service weakens the reference. On the other hand, an additional result or a changed constraint may need revision, which Progressive Commitment handles without assuming that a defined price means unlimited scope.

Paid learning and the next offer

The service can offer a bounded entry point to clarify the intention and produce information: discovery, requirements, prototype and technical analysis can form one commitment with its own outputs, or be sold separately when that helps the buyer. No transaction per operational stage is needed. The initial price pays for the advance assumed, and its outputs support a later offer: what to build, under which conditions, and at which price or consumption rule.

The step between offers does not require complete certainty. The provider decides whether the understanding available allows it to assume a commitment and how to treat what remains uncertain. Requirements, prototype and architecture can support that decision; they are not a formula that reveals the exact cost of construction. Three decisions that can look like one should be kept apart:

Decision	Possible relation to price	What must be preserved
Scope and complexity of the solution	Removing a feature, simplifying a rule or choosing another integration can reduce work and allow a lower offer.	The alternative must stay useful and feasible; dependencies and shared costs prevent a strictly proportional discount.
Depth of discovery	Investigating only what the next decision needs can lower the initial cost, leaving other questions for later.	The commitment must acknowledge residual uncertainty; leaving out needed information can raise exposure, rework and price.
Rigour of acceptance	Checks proportional to scope and consequences are part of the cost of meeting the commitment.	Fewer checks or looser criteria are not the same delivery produced more efficiently.

AI can make it more accessible to produce alternatives with visible economic consequences, so the customer chooses between understandable solutions and commitments instead of assuming that writing less about the same solution makes it cheaper. The next offer must state scope, acceptance, dependencies and conditions for revision, and can contain selectable components as long as the selection remains a feasible delivery. The customer can stop after an accepted discovery; if the entry point only works economically when construction follows, that must be examined across the whole operation, and a buyer’s informed decision to stop is never treated as a failure.

Execution by people, agents and automation

Changing how the work is executed can change cost, time, supervision and risk, and the price can stay tied to the same result where the commitment allows that operational freedom. That freedom must not hide relevant conditions of access, responsibility or evaluation. Automation does not make every saving available for a discount: preparation, evaluation, maintenance and idle capacity can still make up the cost. At the same time, a price based only on the quantity of human work may stop representing a relevant part of a hybrid execution.

Accepted delivery and business benefit

A price tied to a bounded delivery is not a price conditional on business benefit. A feature can meet its conditions and not produce the expected effect because of how it is used, the surrounding process or other dependencies. A component tied to benefit must state which effect will be observed, over what period, and how its relation to the delivery will be examined; a condition based on a poorly observable effect can create disputes instead of alignment. See Outcomes.

Billing, payment and continuation

Price sets economic terms; billing and payment have their own moments and rules. Delivery can consume resources before anything is received, and the agreement must account for that. A payment milestone may coincide with an acceptance, but it need not, and stages must not be redesigned just to produce instalments. Limits, cancellation and remaining obligations must be understandable: stopping can reduce future work without removing costs already incurred, reserved capacity or responsibilities assumed. Internally there may be no commercial price, yet budget and consumption still need tracking, without inventing external revenue to show sustainability.

In the running example

The barbershop’s discovery, priced at 20, has been completed and accepted. The offer to build booking is 80; adding reminders raises it to 95. The customer authorizes booking only, directly or under the additional mandate of 90 described on the Commercial Gate page. The table shows the moment after that authorization and before construction takes place. The values are illustrative and are not internal cost, billing or money received.

Reference	In the example	Meaning
Service already consumed	Discovery completed; contracted price 20	An accepted intermediate delivery; payment follows the agreement.
New commitment assumed	Building booking for 80	Authorized work, still to be done.
Aggregate contracted price	100: 20 plus 80	The sum of both commitments; not everything has been consumed yet.
Additional delegated ceiling	90 for continuing after discovery	The purchasing limit the mandate allows.
Authorization not yet committed	10: 90 minus 80	The balance of the mandate; not a charge or revenue.
Alternative not contracted	Reminders: 15 more; combined build at 95	Exceeds the current additional ceiling and remains unauthorized.
Provider’s internal forecast	Kept separate from the prices shown	Expected consumption can vary without changing the contracted price.

The aggregate authorized exposure for this route would be 110, counting discovery and the additional mandate; the contracted price is 100. None of these figures shows cash received, cost incurred or profit. This visibility can live in a dashboard, a conversation or a simple document; its purpose is to let the buyer understand what has been received, what has been assumed and what is still an option.

Boundaries and failure modes

Every arrangement creates incentives. Charging for effort requires visibility of consumption; pricing by delivery requires care over scope and criteria; benefit-linked components require conditions for observation and attribution. No arrangement removes conflict automatically; clear conditions, good evaluation and the possibility of revision still matter.

The price must be compatible with the operation. Recurring capacity could let a provider serve new demands with less extra preparation, and a delivery-based commitment lets the provider organize those means and keep part of the gains under its own terms. The buyer may benefit through access, time, price or less coordination, but those effects have to be verified. Promising availability or results the operation cannot sustain can make a price attractive and the service unviable; Unit Economics takes that up. The alternatives here are conceptual, not market prices, financial advice or legal models of contracting.

Connected concepts

• Economics overview
• Previous: Commercial Gate
• Next: Unit Economics
• Progressive Estimation, the forecast a price draws on
• Glossary