A large percentage of people I hang out with work in tech. Most of them have only ever known salaried 9-to-5 jobs. Every few months, someone asks me how I ended up doing independent consulting. These conversations have become more frequent in the last year or so. AI-driven (or at least AI-justified) layoffs have tech workers worried about job security for the first time in a while.
None of the advice you’ll read below is something that I came up with on my own. Over the years I’ve come across many online discussions and personal blogs discussing this. Every once in a while you come across something that clicks for you. None of it is groundbreaking or eye-opening either. This advice has barely changed in the last couple of decades. Yet the same questions keep getting asked online and in person, because each new person leaving salaried employment believes their situation is somehow unique. It never is. So here you go.
Why bother
Before getting into the how, let’s spell out what you actually get from going solo.
You own your time. Want more money? Work more. Want a three-week break between engagements? Take it. Nobody needs to approve your PTO. A client making unreasonable demands? Fire them. That last one sounds dramatic, but I’ve done it twice, and both times it freed me up for better work within a month.
Your income isn’t tied to one source. A salaried employee with one employer has a single point of failure. Get laid off and your income doesn’t dip, it vanishes. With multiple clients, losing one hurts but doesn’t sink you. I typically run two or three engagements in parallel, and that matters more than I expected.
This sometimes gets compared to overemployment (OE), where people hold multiple full-time jobs simultaneously. I don’t have anything against it: if you can do your job well while working for multiple employers, good for you. Nobody else will care about providing for your family except you. But independent consulting is a different thing entirely. You’re transparent with clients about what you do and they understand they don’t own 100% of your time. There’s no need to hide the fact that you’re working with other clients (though there’s also no need to go out of your way to bring it up).
You get better at everything faster. When you’re solo, there’s nobody else to handle the stuff you’d rather ignore. You have to get decent at time management, client communication, setting expectations, figuring out which parts of a project actually matter, and saying no to the parts that don’t. A full-time job lets you coast on weak spots for years. Solo work doesn’t provide that luxury, but you get to charge a premium for this.
Enjoy the tax deductions. Your laptop, monitors, software subscriptions, books, conference tickets, flights to industry events, home office space, all deductible as business expenses (rules vary by jurisdiction; I am in Canada).
No corporate politics. No reorgs, no performance review theater. You still deal with organizational dysfunction at client companies, but somehow it feels different when you’re just a tourist in someone else’s mess, and you get to leave when the engagement ends.
You build equity in yourself. Every talk, blog post, and successful engagement builds your reputation, not your employer’s brand. When a salaried employee builds their company’s reputation, the company captures most of that value. When you do it solo, it follows you everywhere.
You can work from anywhere. No commute, no relocation for a job, no asking permission to work from a different city for a week. As long as you deliver and show up to calls, nobody cares where you are.
None of this means solo work is better for everyone. But if you’re on the fence and wondering whether the tradeoffs are worth it, for me the answer has been yes ever since I started.
The network problem
Every veteran consultant says this, and it’s the advice new consultants least want to hear: your first clients come from people you already know. Former colleagues, former managers, that friend-of-a-coworker you met at a conference two years ago. The person you helped debug a production outage at 2am three years ago. That favour is now social capital, and social capital converts to contracts more reliably than any marketing funnel.
Hiring a consultant is a gamble. The buyer is paying real money for an outcome they can’t fully evaluate in advance. A warm introduction from a trusted mutual contact cuts through that risk in a way that no LinkedIn post or cold email can replicate. When I started reaching out to my own network, I was surprised by how many people said something like “I was actually about to start looking for someone who does exactly that.” Many of them actually wanted to hire me full-time, not as a consultant. Sometimes for much more than what I used to make at my last job. That’s where being firm and persistent about your offering and services pays off. You’re a consultant, so act like one.
But before you quit, make a list of every professional contact you’d feel comfortable emailing, not to pitch, but to inform. That list will be your primary source of conversations for the first six months, possibly longer.
The change in mindset
Now let’s talk about psychology a bit. There’s a personality component to consulting that’s worth highlighting. Being genuinely interested in someone else’s problems is both a soft skill to develop and a practical advantage. When a client describes a messy data migration or a team that can’t ship on time, you need to actually care about figuring it out. If you’re faking interest, they can tell, and that’ll show up in the quality of your work later.
The bigger adjustment is how you think about the relationship. A client is not your employer. You’re a peer they brought in to solve a specific problem, and you should act like it. That means pushing back when their plan doesn’t make sense and scoping work honestly instead of saying yes to everything. Most clients respect this, because they’re paying for expertise, not compliance.
The flip side of pushing back is knowing when to stop. If a client hears your recommendation and still says no, let it go. There might be a political situation you’re not seeing, bureaucratic constraints, or a turf war between teams that has nothing to do with the technical merits. You’ll rarely get the full picture. Keep pushing after the first “no” and you become the consultant who doesn’t listen, which is the fastest way to not get your contract renewed. Ultimately, the person who approves your invoices needs to want to keep working with you. That matters more than being right about any single recommendation. (I guess this advice applies equally well to being an employee.)
A related skill that pays off is low-stakes generosity. At conferences, meetups, or in online groups, just listen to people and help out when it doesn’t cost you much. Answer a question, or point someone to a tool that solved the same problem for you last month. Don’t keep score. Some of the people I helped out years ago ended up introducing me to their employer, who later became a consulting client. I couldn’t have predicted or planned for that. It worked because it wasn’t transactional.
The case for narrowing down
A very common mistake is trying to go too broad. “I help companies with software” reads like a Craigslist category. It tells a prospective referrer nothing useful. Compare it with “she does data pipeline work for fintech companies,” which is specific enough to trigger recall. When your former coworker gets asked “do you know anyone who could help with X?”, your name needs to come up immediately, without deliberation. That only happens when your specialization is concrete enough to stick in someone’s memory (it doesn’t hurt if it’s also in high demand).
I resisted narrowing down at first. It felt like I’d be turning away potential work. In practice the opposite happened. A tight focus made conversations shorter and clearer. Prospects understood what I did in a couple of sentences. And the few times someone asked about something outside my niche, I could still say yes if the project made sense.
Pick an industry or a problem type, and commit to it for at least a year. You can always broaden later. Sharpening a vague reputation is extremely difficult once people have already put you in the “generalist” bucket.
Why hourly billing is a trap
Ideally, you want to move away from hourly billing as soon as you can. It punishes competence. If you solve a $50K problem in five hours instead of twenty, you earn 4x less for the same outcome. The better approach is to price based on the value of the result. A client whose team saves twenty hours a week from your data pipeline doesn’t care whether you built it in a weekend or over a month. They care that it works.
Avoid clients who require meticulous time-tracking and justifications for every hour billed. That dynamic turns you into an employee with extra paperwork. Billing “hourly” on paper is fine when both sides agree to a fixed number of hours per month. I’ve had arrangements where the rate is hourly but the expectation is, say, 40 hours a month, and the invoice is the same every month. That’s just a fixed retainer with different labeling. Lawyers, accountants, and PR firms have worked on retainers forever. It’s less common in tech consulting for some reason, but there’s no reason it shouldn’t be. The important distinction is whether the client monitors and polices your time, or trusts you to deliver.
The progression tends to look the same: you start hourly, move to weekly or project-based quotes, and eventually name a price anchored to what the work is worth to the buyer. In other words, you want to be in a position to capture a percentage of the value you deliver, not sell your time by the hour.
Proof of competence, built in advance
Some independent consultants seem to have no trouble finding clients, and they tend to share a common trait: years of accumulated proof, built before they ever planned to go independent. They wrote blog posts because they found the subject interesting, or gave conference talks because they had something specific to say. The audience accumulated as a byproduct.
“Proof” here means anything a stranger can look at and conclude you know what you’re talking about: a blog post about solving a specific technical problem, or a case study with real numbers. The format matters less than the specificity. One thorough post about a real project is worth more than fifty generic AI-slop pieces.
Patrick McKenzie is probably the best-known example. He went from programming to high-end consulting partly by writing in public for years. By the time he was charging $30K a week, prospects already knew his thinking before the first phone call. A handful of solid posts in a well-defined niche will put you ahead of most consultants whose entire public presence is a LinkedIn profile.
If you haven’t started building this kind of evidence yet, start now, even if you have no intention of going solo. Write about what you’re working on. Explain the tradeoffs you’re making and why. This kind of output compounds slowly, but it does compound.
Showing up in the right rooms
Writing blog posts and giving talks are one side of building visibility. The other side is just being present in the places where your kind of work gets discussed.
For me, that’s been ML/AI and MLOps Slack communities, Hacker News, and a few niche forums. I don’t post constantly or try to build a personal brand. I just hang out, answer questions when I have something useful to say, and occasionally link to something I’ve built. Over time, people start recognizing your name. When someone in the group needs help with a problem in your area, or when their company is looking for a contractor, you come to mind.
You can’t predict which conversation will lead to a project six months from now, but you have to be in the room when it happens. The time investment is small (a few minutes a day of skimming and occasionally responding). The compounding effect applies here too.
The free-work trap
People often ask whether offering free work is a good way to land the first client. Free hours signal that you don’t believe your own work is worth paying for. Worse, they attract buyers who are optimizing for cost, which makes them the hardest clients to convert to paid engagements later.
A free thirty-minute intro/diagnostic call is fine. Treat it as a sales tool: use it to show you understand their problem. “I’ll work for free and hope you pay me afterwards” is volunteering. If you want to do unpaid work to build a portfolio, do it for open source projects or nonprofits where there’s no ambiguity about the arrangement. For prospective paying clients, charge from day one, even if the rate is lower than your eventual target.
Saying no
Not every client who can pay is worth taking on. Some projects have warning signs from the first conversation: vague scope, unrealistic timelines, a contact person who can’t make decisions without looping in three other people, that kind of thing. These engagements tend to end with scope creep and late payments.
Early on, you’ll take almost anything because an empty month is terrifying. But after a few bad experiences, you start recognizing the patterns. A prospect who haggles aggressively on a reasonable rate will also fight you on every invoice. How someone negotiates the first contract is an audition for how they’ll behave for the rest of the relationship.
Learning to walk away, sometimes even mid-engagement, is one of the harder skills to develop. But every bad-fit project you decline frees up time for a good one.
Variable income and unstructured time
Finding clients and setting prices are hard, but the thing that catches most people off guard is the psychological weight of variable income and unstructured time.
In a salaried job, money arrives on a predictable schedule. You barely think about it. Working solo, you think about it constantly, especially in the first year. Between invoicing cycles and payment terms, work you finish in January may not reach your bank account until March.
Know your monthly burn rate. Have at least six months of expenses saved before you start, twelve if you can manage it. Track your pipeline and your cash position separately. And assume that the first year will be financially worse than your salaried job, because chances are it will be. The second year is usually a different story.
The time-management problem is subtler. Nobody assigns you work. If you’re used to sprints, standups, and a manager setting priorities, the sudden absence of structure can be paralyzing. I found that dividing my week into rough blocks (e.g. client work in the mornings, business development and admin in the afternoons) was enough scaffolding to stay productive without recreating the corporate calendar I had just left.
The compounding effect
I want to talk about The Compounding Effect yet again, just because how important it is in general, and in consulting specifically. When you work across multiple clients and project types, your skills improve at what feels like 3-5x the rate of a single full-time job. You’re not stuck in one codebase with one team’s assumptions. You see the same class of problem in different contexts, and each occurrence improves your ability to solve it.
This has two practical effects. First, you become much more productive. A data pipeline migration that would have taken you three weeks in year one takes you one week in year three, because you’ve done four of them for four different companies by then.
Second, clients genuinely want the outsider perspective. They’re stuck inside their own organization, and they know it. When you can say “I’ve seen three other companies try this approach, and here’s what actually worked,” that carries weight that no internal engineer can match.
It takes a couple of years to get this loop started, but once it’s running, finding work gets noticeably easier.