Skip to content
Branemind
All insights
White paperCommercial model··12 min read

Stop billing for effort. Start billing for outcomes.

A delivery-based commercial model for AI, data and digital engineering, plus an honest account of exactly where outcome accountability stops.

Billing unit
Accepted delivery
not hours
Framework
6 steps
define → measure
Accountability
3 layers
committed / joint / external

Measured against a baseline captured before the work started. One engagement, not a forecast for another business. See the methodology note below.

Technology services have traditionally been sold through people: a number of engineers, a rate card, and a monthly calculation of hours consumed. That model is easy to administer, but it can disconnect commercial incentives from the reason a client engaged a technology partner in the first place: to get something delivered.

Branemind uses a delivery-based billing model. We define the intended deliverable, establish measurable acceptance criteria, agree commercial milestones, and bill against accepted delivery. The unit of commercial conversation therefore shifts from 'How many people did we deploy?' to 'What capability or outcome did we deliver?'

This does not mean every business outcome can be guaranteed by a technology provider. Revenue, adoption, regulatory approval and operational performance can depend on factors outside a delivery partner's control. A credible delivery-based model distinguishes between what the partner can commit to, what can be jointly measured, and what remains a client-side dependency.

The commercial misalignment in headcount billing

In a headcount model, the primary billable unit is capacity. More people or more time generally creates more billable value for the supplier, even when the client's objective is to reduce cycle time, automate work, simplify architecture, or require fewer interventions.

That is a structural tension: efficiency benefits the client, but may reduce the supplier's billable base. Delivery-based billing attempts to align both sides around completion, quality and usefulness instead.

01Effort

People, hours, timesheets.

Headcount bills here
02Output

Code, pipelines, tickets closed.

03Outcome

A bounded capability that passes acceptance.

Delivery bills here
04Value

The business change that follows.

The same engagement, four places a price could attach. Headcount billing attaches at the first link. Delivery billing attaches at the third, close enough to the client's intent to be meaningful and far enough back to still be something a partner can be held to.

Headcount vs delivery: a different unit of value

Primary unit
Headcount-basedPeople / hours
Delivery-basedAccepted deliverable / milestone
Client asks
Headcount-basedWho is allocated?
Delivery-basedWhat will be delivered?
Supplier incentive
Headcount-basedUtilisation
Delivery-basedCompletion and quality
Scope control
Headcount-basedOften managed through capacity
Delivery-basedManaged through explicit outcomes and change control
Measurement
Headcount-basedTimesheets / effort
Delivery-basedAcceptance criteria / KPIs
Efficiency
Headcount-basedCan reduce billable hours
Delivery-basedCan improve delivery economics
Risk allocation
Headcount-basedMore delivery risk remains with client
Delivery-basedMore execution accountability sits with partner

What Branemind means by 'delivery'

A delivery is not a vague promise such as 'improve AI' or 'modernise data'. It is a bounded, testable capability with a clear definition of done. Depending on the engagement it can be a production workflow, an integration, an agentic process, a migrated data product, a reconciliation capability, a measurable automation milestone, or another agreed unit of business functionality.

The strongest deliverables combine four elements:

  • A business objective: what this is for, in the client's language.
  • A technical artifact or capability: the thing that actually gets built.
  • Acceptance criteria: observable and testable, agreed before implementation begins.
  • Evidence that the criteria were met: the artefact that triggers the milestone.

The outcome contracting framework

  1. 1Define the outcome
  2. 2Establish a baseline
  3. 3Agree acceptance criteria
  4. 4Build and validate
  5. 5Accept the delivery
  6. 6Measure impact

Step 6 feeds step 1 of the next phase. Impact informs what gets scoped next, without pretending every external variable is the delivery partner’s to control.

Six steps, run in order. The discipline is in steps 2 and 3: a baseline captured before the work starts, and criteria agreed before anyone writes code.
  1. Define the outcome

    Translate the business problem into a bounded result. Avoid using team size as the definition of scope.

  2. Establish a baseline

    Where measurement matters, capture the current state: cycle time, manual touches, accuracy, failure rate, throughput, cost per transaction, or another relevant metric.

  3. Agree acceptance criteria

    Define what evidence constitutes completion. Criteria should be observable, testable and understood before implementation begins.

  4. Build and validate

    Branemind chooses the engineering approach, automation level and internal resourcing needed to reach the agreed delivery.

  5. Accept the delivery

    Client and Branemind review the evidence against the agreed criteria. Commercial milestones are connected to this acceptance.

  6. Measure impact

    Where appropriate, track operational or business KPIs after release. Impact metrics inform subsequent phases without pretending every external variable is under the delivery partner's control.

How tangible outcomes can be quantified

Delivery-based billing works only when 'done' can be made concrete. The measurement method should match the nature of the work.

Work typeExample deliverablePossible acceptance evidencePossible impact metric
AI agentProduction-ready workflowTool-call tests, guardrails, eval threshold, deploymentAutomation rate / handling time
Data engineeringReliable data pipelineFreshness SLA, reconciliation checks, observabilityLatency / incident reduction
IntegrationSystem-to-system workflowSuccessful test cases, idempotency, error handlingManual steps removed
Process automationAutomated operational processDefined scenarios complete without manual executionCycle-time reduction
AnalyticsDecision-ready data productMetric definitions, validated outputs, access controlsTime-to-insight / adoption
Acceptance evidence is what triggers a milestone. Impact metrics are tracked afterwards and inform the next phase. They are not the same commitment.

A simple commercial example

Consider a reconciliation process that currently requires repeated manual matching across invoices, ledgers and bank transactions. A headcount contract may sell two engineers for a period of time. A delivery contract instead defines the capability: ingest agreed sources, match transactions according to approved rules, surface exceptions, preserve an audit trail, and meet agreed validation thresholds.

If Branemind can reach the agreed result through better tooling, reusable components, automation or a smaller expert team, that efficiency is not penalised. It becomes part of the value proposition.

Why this matters more in the AI era

AI-assisted engineering changes the relationship between effort and output. Code generation, reusable agent patterns, automated testing, infrastructure automation and increasingly capable models can compress work that previously required larger teams. If commercial models remain tied only to hours, clients may pay for an old proxy even as the underlying productivity curve changes.

AI-leveragedHeadcountEffort (people × time) →Delivered output →

Illustrative shapes, not measured data. The point is the direction, not the gradient.

Under a headcount model, output is assumed to track effort in a straight line, which is precisely what an hourly rate card prices. Engineering leverage bends that line, and an hours-based contract cannot see the difference.

Delivery-based billing makes the commercial model more compatible with AI-enabled productivity. The provider is rewarded for engineering leverage, while the client retains focus on the capability received.

What this changes for the client

  • Budget conversations become connected to defined milestones rather than open-ended utilisation.
  • Procurement can compare proposals on scope, acceptance and value, not only rate cards.
  • Engineering partners have an incentive to automate their own delivery process.
  • Smaller expert teams can compete on capability instead of staffing volume.
  • Governance becomes clearer, because change requests can be separated from the agreed definition of done.

What it requires from the delivery partner

This model is not simply fixed-price billing with a new label. It requires stronger discovery, sharper scoping, reusable engineering assets, evaluation discipline, observability, documentation, risk management and transparent change control. A supplier that cannot define and measure delivery will struggle to price delivery responsibly.

Where delivery-based billing needs guardrails

Not every engagement is equally suited to a pure outcome model. Early-stage research, ambiguous discovery, rapidly changing requirements, third-party dependencies and client-controlled adoption can make a final business outcome impossible to price responsibly. Branemind's approach is to make accountability explicit rather than over-promise, by separating three layers.

Committed deliveryTied directly to acceptance

A tested integration deployed to the agreed environment

Joint outcomeMeasured with shared dependencies

Reduction in manual handling after client rollout

External business resultTracked, but not falsely guaranteed

Revenue growth, regulatory approval, end-user adoption

Accountability is strongest at the top and deliberately weaker further out. Naming which layer a metric belongs to is what keeps the model honest.

The three layers, and the commercial treatment each one can honestly carry. Most disputes come from treating the third layer as though it were the first.

Delivery-based does not mean unlimited scope either. If the target outcome, source systems, acceptance criteria, compliance requirements or dependencies change materially, the commercial scope should change transparently as well. This protects both the client and the delivery partner.

Branemind's position

Branemind believes the technology services industry should progressively move from selling capacity to selling accountable delivery. Headcount will remain relevant for some forms of staff augmentation and open-ended programmes. But when a client hires a specialist partner to solve a defined problem, the commercial model should reflect the problem solved.

We do not want clients asking how many hours Branemind spent. We want them asking what Branemind delivered, and what changed because of it.
Branemind

A call to buyers and builders

  • For buyers: ask technology partners to make the definition of done visible.
  • For builders: invest in the engineering systems that let you accept delivery accountability.
  • For procurement teams: evaluate commercial models not only by day rate, but by the clarity of the outcome, the quality of the acceptance criteria, and the risk each party is prepared to own.

Methodology and privacy

  • Clients are described rather than named. Where a quote appears it is attributed to a role and published with approval.
  • Figures come from a single engagement and are not a forecast of what a different business would see. Volume, data quality and process maturity move them more than the technology does.
  • Baselines were captured before the work started. The measurement window and sample for any figure on this page are available on request.
  • Where a number is illustrative rather than measured, it is labelled as such in the text.
If this looks like your loop

Review your path to production

A readiness read on evaluations, controls and ownership.