How to Sell Small, Fixed-Scope Consulting Offers

A prospective client is interested in what you do. They are not ready to sign a six-month programme with someone they met forty minutes ago. Both things are true at once, and most proposals ignore the second. The usual result is a polite delay that never ends.

This article is about the alternative: small engagements with a fixed scope, a short calendar and one deliverable. My own work is AI consulting, so the examples lean that way, but the shape applies more widely. The quotations come from my book The Micro-Consulting Playbook.

Why a small offer gets a yes

A small engagement is small in shape, not in ambition. It has a fixed scope, a short calendar and one deliverable the client can point at when it is finished.

A small yes is still a yes. It is also the only kind a stranger can give you.

It works for plain reasons. A two-week piece of work is short enough that nobody needs board approval, a procurement process or a steering group. The buyer can explain it to a colleague in a sentence. If it goes badly, the loss is limited and everyone can see where it ends.

It also suits the way the work is now done. Hourly billing has a quiet flaw that gets louder once you work with AI tools: it punishes you for getting faster. Every hour the tooling saves is an hour you cannot invoice. A fixed fee for a defined result removes that problem. The client pays for the outcome and your speed becomes your business.

One habit comes before any offer. Treat the first call as a diagnosis and not as a pitch. Most consultants spend it explaining what they do, which the buyer could have read on the website. The better use is finding out what is wrong, which only the buyer can tell you.

Three shapes worth knowing

You do not need many offers. A few shapes cover most situations.

The paid audit. A fixed-scope review of how a team works today, ending in a roadmap. It is low risk for the buyer, paid work for you, and short enough to finish before anyone loses interest. It must end in a decision. An audit that produces a document and no decision has not done its job.

The two-week sprint. One problem, one working result. The fortnight is a container more than a deadline: it stops the work from spreading. A common version is the workflow fix, where one painful, recurring process is diagnosed, fixed, verified and handed over.

The pilot. A larger test of whether something can run inside the client's business. Treat it as an experiment with a hypothesis, a method and a result. Without those it is a demo with a longer invoice.

Arranged in order, these form a ladder. At the bottom is something free and small: a useful article, a short diagnostic call, a specific analysis of the prospect's own problem sent before any meeting. Above it is the audit, which finds the problem. Above that is the build that fixes it, and above that a retainer that keeps it working. Each step makes the next one easier to agree.

Fix one thing properly and the client will show you the other nine.

Name it and narrow it

An unnamed service is a favour. A named one is a product with a price. People file things by name, and the person you are talking to usually has to explain the purchase to someone else.

If they cannot say it to their boss, they cannot buy it from you.

Narrowing matters for the same reason. Generalists compete with everyone. A specialist in one sector meets the same problems again and again, and small engagements depend on that repetition. The third time you build the same thing, it has stopped being a project and become a product. Write the process down. A process you can describe is one you can improve, hand to a tool and deliver in less time.

Deliver without drama

Fixed scope only works if delivery is orderly. Five practices do most of the work.

Get access first. Nothing ships until you have credentials, access to the data or repository, and a named decision-maker. On a two-week engagement, three days spent waiting for a password is a fifth of the calendar gone.

Show work every week. A weekly demo, working and on screen with the client's kind of data, takes about half an hour. It is cheap insurance against building the wrong thing, and it means bad news arrives in small pieces.

Agree how acceptance is measured. "Are you happy?" depends on mood and on the last thing the client saw. A scored set of test cases, agreed in advance, turns sign-off into a result. If a case fails, there is something specific to fix.

Have a protocol for extra requests. Scope creep rarely arrives as a demand. It arrives as an enthusiastic, reasonable, small-sounding idea from a client who is pleased with the work. Say yes to the idea and no to the timeline. Write it on a list of parked requests. That list is also your next proposal, written for you by the client.

Hand over properly. Write a runbook so the result can be operated by someone who has never met you. Make sure one person inside the client can demonstrate it without you. Hand over the prompts and configuration as well. Keeping them back to create dependency is a short game that ends badly.

Then close the loop. Measure what changed and tell the buyer in their own terms. Work that nobody measured tends not to be renewed or referred.

Small engagements end quickly, so use the finish to set up the next one. Write the case study while the numbers are fresh and the client is pleased: problem, approach, result and a line from the client. Ask for a referral at that same moment, and ask for something specific, because a vague request to "keep me in mind" produces nothing. Ten small engagements earning steadily are a sturdier business than one large bet.

The shape of the price

This section has no numbers in it, on purpose. Figures depend on your market. The structure does not.

Anchor on the problem. Every engagement has two costs: what the problem costs the client over a year or more, and what the work costs you. Price against the first.

Ask about budget before you quote. It feels awkward the first few times. The answer is often higher than the figure you were about to say. When it is lower, you have learned that before writing a proposal.

Offer three options. One price invites a yes or a no. Three options invite a choice between them, which is a better question for both sides.

Charge for discovery. If you spend days investigating a problem before any agreement, you are giving away the most valuable part of the work. That is the reason to make the audit a paid product.

Take a deposit before work starts. It is standard practice and it shows that both sides have decided.

Separate the build from the care. Charge properly for the build, then charge a monthly fee to keep it working. Most systems need looking after as the business changes.

Know your running costs. If delivery involves paid model calls, each one costs something. A fixed price with an unknown cost is a bet you did not mean to place.

What to take from this

• A stranger can agree to something small. Give the engagement a fixed scope, a short calendar and one deliverable.
• Use the first call to diagnose. Pitch later, if at all.
• Three shapes cover most needs: the paid audit, the two-week sprint and the pilot. Arrange them as a ladder.
• Name the offer so the buyer can repeat it to their boss.
• Secure access first, demo weekly, agree a scored acceptance test and park extra requests in writing.
• Price against the cost of the problem, offer three options, charge for discovery and take a deposit.
• Hand over everything, measure the result and write the case study before you leave.

If you want the full version, The Micro-Consulting Playbook: Small AI engagements that clients say yes to shows how to scope, shape, deliver and prove small engagements built with Claude, from the paid audit to the agent pilot, in 100 short chapters with a labelled diagram in each. It discusses how to structure a fee and does not tell you what you will earn. It is a 314-page PDF and costs £9 on Gumroad: https://matsiems.com/books/the-micro-consulting-playbook.