Fractional, consulting or software: which business should you start after tech?

If you need income within a few months, start with fractional or consulting work. Build software when you can go a year or more without it paying you, or when client work already covers your bills while you build. That's my rule, and I run both kinds of business myself.

What business models can a tech leader start after a corporate career?

You have six realistic options: a fractional role, consulting or advisory, a productized service, an agency, software (often micro-SaaS) and an audience or media business. They differ in how fast they pay, what they cost to start and what your week looks like.

You know most of these already, so here's the one-line version of each, so we mean the same thing:

• A fractional executive holds an ongoing part-time role at one company, like head of growth two days a week, usually on a monthly retainer.
• Consulting or advisory is paid judgment on a defined problem. You sell it as a fixed-fee project or a monthly retainer for access to you.
• A productized service is a service sold like a product: same scope, price, timeline and output for every client.
• An agency is a services business where a team does the client work and you run sales and people.
• Micro-SaaS is a small software business, usually self-funded and run by one to three people, selling a subscription to a narrow niche.
• An audience or media business makes money from attention: a newsletter, a community, courses or sponsorships.

You can combine them, and I get to combinations further down. Look at them one at a time first.

How do fractional, consulting and SaaS compare?

Fractional and consulting work are the fastest and cheapest to start. Software and audience businesses are the slowest to pay, and the only ones with no hard ceiling on income. Everything else sits in between.

The time ranges below are my own rough estimates for someone with a decent professional network. Read them as orders of magnitude.

How do I choose between a services business and a software company?

Decide on five things, in this order: how long you can go without income, how much risk you can carry, how far past your own hours the income needs to grow, which skills you're willing to learn and what you want a normal week to look like. The first two rule options out. The last three pick between what's left.

1. Time to first revenue: how long can you go without income?

Your runway is your savings divided by your monthly costs, and it matters more than your preferences. My rule: with less than a year of runway, start with a model that pays in weeks.

Services pay fastest because you're selling work you already did in a job, to people who watched you do it. A former VP of product can sell a pricing review to a founder she knows within a month of deciding to.

Software is slower than it looks from the outside. Stripe's 2025 review of Atlas startups put the median time from incorporation to first payment at 34 days. The first payment is quick. Getting to a salary is the long part. ChartMogul's 2025 SaaS Growth Report (6,525 software companies) found that 3.3% reached $1M in annual recurring revenue within a year of first charging customers, 13.4% within three years and 25.1% within five.

Now do the salary maths. Say you want 10,000€ a month before tax. Illustrative numbers:

• Fractional: two clients at 5,000€ a month each.
• Productized service: four fixed-price audits at 2,500€ each, every month.
• Micro-SaaS at 49€ a month: 205 paying customers, kept after churn.

Two clients means a few good conversations with people who already trust you. Two hundred and five subscribers means you've built a marketing system, and that takes a lot longer.

2. Capital and risk: what are you putting on the line?

Services put very little cash at risk. Software puts your time at risk first, then your money. Every model has its own way of going wrong, and I'd name yours before you pick.

• Fractional and consulting: concentration. With two clients, losing one halves your income overnight.
• Agency: payroll. You commit to people before the next retainer renews.
• Software: months of building for demand you haven't confirmed. Hire a developer and it's your cash too.
• Audience: a year of publishing with no guarantee readers will buy anything.

If you're still employed, the picture changes. Software and publishing fit around a job, because evenings work for both. Consulting and fractional work need you in meetings during business hours, which is hard to hide in a full-time calendar. Whatever you pick, read your employment contract for outside-work and intellectual property clauses before you start. Do it twice for software.

3. Scale: how far past your hours does income need to grow?

This is how much more you can earn without working more hours, and services have the least room. You can calculate a fractional or consulting ceiling in one line.

Say you charge 1,000€ a day and can bill 12 days a month, because selling and admin eat the rest. Your ceiling is 12,000€ a month. You raise it in two ways. Charge for value instead of time, which is the core of Alan Weiss's and Blair Enns's work on pricing expertise. Or narrow what you're expert in, which David C. Baker argues is what lets specialists charge more. I'd do both. My guide on how to price fractional and consulting work walks through each.

Productized services lift the ceiling because you deliver faster every time you repeat the work. Software lifts it most, since one more customer costs you very little to serve. Every customer still costs money to win, though, and that cost never shrinks on its own.

4. Skills: what will you have to get good at?

Every model hands you one job you probably never did in your corporate role. Pick the one you're most willing to learn, because you'll be doing it every week.

• Services: selling yourself in a conversation and saying no to work outside your scope. Blair Enns's Win Without Pitching Manifesto is the best short read on this.
• Productized services: packaging. Alex Hormozi's $100M Offers is useful on structuring scope and guarantees so a fixed offer sells.
• Software: distribution and positioning. April Dunford's Obviously Awesome covers positioning, Bob Moesta's work on jobs to be done explains why people switch products, and Madhavan Ramanujam's Monetizing Innovation makes the case for testing price before you build.
• Agency: hiring, and writing delivery down so other people can run it.
• Audience: publishing on a schedule, for a long time, with opinions.

Engineering leaders can build the product. In a small software company, getting it in front of buyers is the harder job, and it lands on you whether you like it or not. My advice: plan the selling before you plan the build.

5. Lifestyle: what do you want an ordinary Tuesday to look like?

Picture a normal working day in each model and notice which one you'd look forward to. A fractional Tuesday is back-to-back calls and a team channel that expects replies. A software Tuesday is quiet building, a support ticket and a marketing task you've avoided for a week. An agency Tuesday is mostly other people's questions.

People pick a business model for its best month. You'll spend most of your time in its ordinary weeks, so pick the week.

Is fractional work a real business or a job with more bosses?

Fractional work is a business once you have a named offer, more than one client and a repeatable way to find the next one. Without all three, it's a part-time job with less security.

The warning signs are easy to spot. Every project starts from a blank proposal. Your rate hasn't moved since you started. All your clients came through one former colleague. The fixes are just as clear: write one offer with a fixed shape, cap how many clients you take at once, raise your price when your calendar fills and give people a way to find you that doesn't depend on luck. My guide on how to go fractional covers the setup step by step.

Should I start with services and build software later?

Yes, if you have less than a year of runway. Services pay the bills, and client work shows you which problem is worth turning into software. The risk is that client work fills every hour and the product never starts.

Rob Walling's stair step approach is the best-known version of this. You start with a simple one-time product sold through one channel, repeat until it's reliable, then move to recurring software. For ex tech leaders I'd tweak it:

1. Services. Sell your judgment to a specific buyer with a specific problem.
2. Productized service. Once you've solved the same problem three or four times, fix the scope and publish the price.
3. Software. Automate the part of the productized service you repeat most, and sell it to the same buyers.

I'd hold myself to three rules. Keep the same buyer at every step, so your network and reputation carry over. Block build time on your calendar before clients can take it, like one fixed day a week. And write down the trigger for moving your time, for example product revenue covering your fixed costs three months in a row.

Some hybrids are stable on their own and never need to become software:

• A fractional role plus a fixed-price audit that brings in the next fractional client.
• An audience plus a small advisory practice fed by readers who reach out.
• Consulting plus a micro-SaaS tool sold to the same buyers.

My own portfolio mixes both sides: Madgrowth sells advisory and digital products, and Mazo is a software company. You can run the two side by side. It only works with the calendar discipline above.

Which business model fits your Builder Archetype?

Your Builder Archetype describes how you like to work. Each of the five has models that suit it and models that will wear it down, and I'd take the second list as seriously as the first.

• The Advisor fits fractional leadership, advisory retainers and a fixed-price teardown. Few clients, each paying properly for judgment. The trap is underpricing until you've rebuilt a job.
• The Productizer fits productized services, templates and self-serve courses. Revenue grows with copies sold. The trap is shipping several small things and properly selling none of them.
• The Venture Builder fits micro-SaaS and AI tools in a niche you already know, often funded by consulting at the start. The trap is building for another quarter before talking to a customer.
• The Orchestrator fits a boutique agency or a done-for-you service run by a small team on your playbook. The trap is building delivery for ten clients while the pipeline holds one.
• The Broadcaster fits audience-first businesses: a newsletter, cohort programmes and launches, with advisory work coming from readers. The trap is growing reach with nothing packaged to sell.

Most people have a main archetype and a second pull. An Advisor with some Venture Builder in them might do fractional work for two years, then build a tool for the problem they kept fixing. That's the stair step from the previous section, shaped around how you work.

Questions people ask