An independent consultant is someone who sells their own expertise straight to the client, by the day, the project or the month, with no employer in between. Setting one up takes four decisions, and the paperwork is the smallest of them: decide what you are not, build a graded rate card from a borrowed ladder rather than a guessed number, write a page of terms, and have ten conversations with people who already have the problem you solve. Everything else, including the company registration and the website, can wait for a quiet afternoon. What you cannot postpone is the arithmetic: how many days you can bill, which is your utilisation (utilization, in the American surveys), and what each of those days must earn to carry the days you cannot.
I have started twice. In 2013 my co-founder Steve and I began the business that became DevOpsGroup, a services company helping other businesses build and run software, with a blog, a questionnaire and no customers. I now run Critical Cloud, a services business that helps companies keep their software and cloud systems running reliably. What follows is the order I would use if I were setting up alone this month.
Setting up: the paperwork is the easy part
In the UK you will choose between a limited company and trading as a sole trader. I am not an accountant, so make that choice with one; the answer depends on how much you expect to earn, who your clients are and how they prefer to contract. Whichever you choose, you need a business bank account, professional indemnity insurance at a level your first client’s procurement team will accept, an invoice template, and a one-page terms document covering payment terms, expenses, cancellation and intellectual property.
That is a week of admin. Do not let it become a month. In 2013 we had been talking to prospective customers for weeks before we incorporated, and the conversations taught us more than the paperwork did. Do not buy a logo. Do not build a website until you have a sentence you would be happy for a competent stranger to argue with, because the website will only repeat whatever sentence you have.
The rate card with no billing history
You cannot value-price without a record of delivered value, and you cannot average deals you have not done. So the first move is calibration, not invention: borrow a ladder that already exists. Public procurement frameworks publish graded day rates for most kinds of professional work, and large firms’ rate cards leak constantly. Find the ladder for your discipline and place yourself on the rung that matches the work you will actually do, not the rung that flatters you.
Then do three things a single guessed day rate never does.
Define the day. How many hours, whether it is on site or remote, what happens to travel time and expenses, and what a half day costs. Most disputes about consulting fees are really disputes about what a day was.
Put the terms on the card. Payment terms, minimum booking, cancellation notice and the date the rates are reviewed belong on the same page as the numbers. A price without terms is an invitation to negotiate the terms later, when you have less leverage.
Publish a graded card even though you are one person. This sounds odd. It works. In 2014 we sent a buyer a five-grade card when the company was two founders and a handful of people we could call on, and we later learned the buyer had priced their bid on the middle rung. Nobody told them to. Publish one price and a buyer negotiates it. Publish a ladder and a buyer chooses the rung that looks like the work, which is rarely the cheapest and rarely the dearest. The ladder also tells you, later, who to hire first: the rung where the revenue is.
Model your income on the rung you will actually sell, not on the average of the ladder. If you need help with the arithmetic from cost base to billable days to a rate you can defend, the guide to setting a consultant day rate in the UK walks through it, and the consulting rate card playbook in development covers the borrowed ladder, the day definition and the blended rate in full.
One rule above all the others. When a buyer says the proposal is too expensive, change the work, not the rate. In January 2014 a prospect told us they liked the idea and could not afford it, and asked what they could do on their side. I spent a night not offering the obvious discount. What went back was a smaller scope, a minimum monthly commitment and a break clause, at the same rate. Steve’s reply, as I remember it, was six words: “Rate is fixed, effort is flexed.”” It stayed the house answer for as long as we owned the company. Every rate you concede becomes the rate that client expects for ever.
The first client
The first client almost always comes from people who already know your work: a former employer (read your contract first), former colleagues who have moved on, the vendor whose product you know well enough to be introduced by, or the user group where you have been answering questions for years. Strangers arrive in year two.
Expect the first paid work to be adjacent to what you set out to do. Ours was an interactive infographic for a digital marketing agency, which was not in any plan we had written. We quoted it two ways, a flexible range for the effort used and a fixed price against an agreed brief, and the agency chose the fixed price. Take adjacent work if it pays and teaches you something about how buyers choose. Do not let it define you.
Where you can, sell a diagnosis before the cure: a short, paid piece of work with a question it answers, a deliverable it leaves behind and a point at which the client can stop. It is easier to buy than a programme, and it earns you the right to propose the programme. Before you pitch anything, ask about the last time the problem happened, how often it happens, what it cost, what they have already tried, and what would convince them it had improved. Those five answers write the proposal for you.
The two-week positioning test
Most new consultants describe what they do in the vocabulary of their category, and the category word has stopped meaning anything in a sales conversation. The correction is to write down what you are not. Find the laziest common reading of your category, then write the sentence a competent stranger could actually argue with.
Send it to between five and ten practitioners chosen because their disagreement would be worth having, not because they might buy. Give it two weeks. Silence tells you the sentence is too safe. Polite agreement tells you it is too vague. Argument from someone competent is the first real market signal you will get, and it costs nothing but the fortnight. That was how we started in 2013, before we had a company number: a blog with a point of view, and a small number of people we hoped would push back. The consulting niche playbook, in development, sets out the seven steps and the three tests.
What to watch in the first year
Utilisation. You will bill fewer days than you planned, because selling, administration and the gaps between engagements are not billable, and the rate has to carry them. Cash. Your payment terms are your cash flow forecast; thirty days on the card becomes longer in practice, so invoice on the day the milestone lands. And the record. Keep a note of every time the rate held and the scope moved. A year from now it will be the only thing that persuades you, under pressure, that the number survives pressure.