Before the conversation

What an engagement costs.

Most engineering firms answer this with a rate card or with silence. Neither helps a buyer who has to defend a number internally before anyone has spoken to a supplier.

What follows is the list of things that actually move the cost of an engagement, what is needed to produce a figure, and what would change that figure once work has started.

Build your brief
DRIVEN BYScope and unknowns
AGREEDBefore work starts
REPRICEDIn writing, not quietly

01 / WHAT MOVES THE NUMBER

Cost follows uncertainty, not ambition.

How much is already known

An outcome that is specified, owned and testable costs less to deliver than the same outcome described as a direction. The single largest cost driver is not the size of the build; it is how much has to be discovered while building it.

How many systems it touches

Every integration adds an owner, a credential path, a failure mode and a test. Two integrations are not twice one. The count matters more than the size of any individual system.

What the data is really like

Migration cost is set by how far the data has drifted from what its schema claims, not by its volume. This is routinely discovered after pricing, which is why it is looked at before.

What evidence the result must produce

Work that has to satisfy an auditor, a regulator or a customer security review carries the cost of producing evidence, not just the cost of being correct. That is real work and it is scoped as work rather than absorbed.

Whether the environment can be reproduced

If the system cannot be stood up outside production, every change costs more and carries more risk. Where that is the situation, making it reproducible is usually the first thing worth paying for.

Who has to agree

Decision latency is a cost. An engagement with one accountable owner moves at a different rate from one requiring a committee, and the difference shows up in duration rather than in day rate.

02 / HOW A NUMBER IS PRODUCED

Assess, then scope, then price.

01 · Brief

The outcome, the constraints and the deadline that actually matters. This costs nothing and can be produced on this site without contacting anyone.

02 · Assessment

Separately scoped work that establishes the current state: what exists, what it integrates with, what the data is like, and what would have to be true for the outcome to be reachable. It ends in a written position, not a proposal.

03 · Scope and price

A written scope with stated exit criteria, priced against that scope. Commercial terms, responsibilities and fees are agreed in signed engagement documents before delivery starts.

04 · Staged delivery

Work is delivered in stages with exit criteria, so cost is committed in increments and the engagement can be stopped at a stage boundary rather than only at the end.

03 / WHAT CHANGES A PRICE

Repricing happens in writing, before it is built.

Scope change

Anything added to the agreed scope is recorded and repriced before it is built. The alternative — absorbing it and recovering the cost through a slower schedule or a thinner test suite — is worse for both sides and harder to see.

Discovery that invalidates an assumption

Where the assessment recorded an assumption and reality contradicts it, that is stated as it is found rather than at the end. An assumption written down before work starts is what makes this a conversation instead of a dispute.

Dependencies outside the engagement

Waiting on a third party, an access approval or another team is not billed as delivery, but it does move dates. Where a dependency is on the critical path it is named in the scope so the risk sits where it actually is.

04 / WHAT IS NOT PUBLISHED

The boundary, stated plainly.

No rates or ranges appear here

Not because they are secret, but because a number without a scope is not information. Any figure quoted before the systems are understood would be a guess presented with more confidence than it deserves.

This page is not an offer

Descriptions here are informational. Commercial scope, fees, service levels, warranties and data-processing terms are agreed in signed engagement documents, as set out in the website terms.

05 / QUESTIONS

About cost.

Why is there no rate card?

Because the same deliverable costs different amounts depending on what is already known. A rate card that covered the worst case would be padded for everyone, and one that covered the best case would be revised on contact. What is published instead is the list of things that move the number, so a budget can be formed before a conversation rather than during one.

What is needed to produce an estimate?

The outcome being bought, the systems it has to work with, who owns the decisions, and what evidence the result has to produce. Access to the running system matters more than documentation about it. Where those are not yet known, the honest first step is a scoped assessment that establishes them.

Is the cost agreed before work starts?

Yes. Commercial scope, responsibilities and fees are agreed in signed engagement documents before delivery begins, as set out in the website terms. Work that has not been scoped and agreed is not started on the assumption it will be covered.

What makes an estimate wrong?

Unknowns that were not surfaced before pricing: an integration nobody owns, data that does not match its schema, an environment that cannot be reproduced, or an approval step discovered late. This is why the assessment exists as separately scoped work rather than as free pre-sales effort.

Does a fixed price mean a fixed scope?

It has to. A price is fixed against a written scope with stated exit criteria. If the scope changes, the change is recorded and repriced before it is built, rather than absorbed quietly and recovered somewhere less visible.

Bring the constraints, not the specification.

The brief takes a few minutes, stays in your browser, and gives a cost conversation something concrete to start from.

Build your brief