Most side businesses fail at the same point: three months in, having built a website, a logo, a company and a social presence, with no customers and no evidence anybody wants the thing.
The businesses that work do the opposite. They sell something in the first month, badly and manually, then build the machinery around whatever people actually paid for. This is that route, laid out for somebody with a job and ten hours a week.
Quick Info
Hours needed
Eight to twelve a week, consistently
First milestone
One paying customer within six weeks
What to build first
Nothing. Sell manually and build afterwards
Legal minimum in the UK
Register as self-employed with HMRC
What to check
Your employment contract
When to go full-time
When it replaces 60% to 70% of your salary
Check your employment contract first
Before anything else. Many contracts contain clauses on outside work, intellectual property and non-competition, and discovering them after you have customers is a bad time to find out.
—Look for exclusivity, outside interests and IP assignment clauses.
—Anything you create on employer time or equipment may belong to them. Use your own laptop.
—Do not compete with your employer or approach their clients. This is the fastest route to a legal problem.
—If the contract requires permission, ask. Most employers agree to unrelated side work.
—Take proper advice if the clauses are ambiguous and the business is significant.
Pick something you can sell within a month
Not the most exciting idea. The one where somebody could pay you in three weeks, because early revenue is the only thing that tells you whether any of it is real.
—Services beat products. You can sell a service before it exists; a product has to be built first.
—Sell a skill you already have. Learning the skill and finding the customers at once is two businesses.
—Pick a customer you already have access to. Your industry, your former colleagues, your local area.
—Prefer businesses to consumers for a side business. They pay more, pay faster and buy repeatedly.
—Avoid anything needing stock, premises or upfront investment until you have proven demand.
The idea you can test in three weeks is worth more than the idea you love and cannot start for six months.
Validate by selling, not by asking
Surveys and "would you use this" conversations produce encouragement, not evidence. The only validation is somebody paying.
01Write down exactly what you are offering, who for, and the price.
02Tell twenty people who fit that description. Individually, directly, by message or phone.
03Offer it to them at a real price. Not free, not discounted to nothing.
04If two say yes, you have a business to build. If nobody does, change the offer or the audience, not the marketing.
05Deliver it manually and badly. Systems come later.
This is uncomfortable and it is the whole thing. Everything else on this list is downstream of somebody paying you once.
The legal and tax minimum
In the UK, and check the specifics for your situation because I am not an accountant.
01Register as self-employed with HMRC once you earn above the trading allowance. This is a form, not a process.
02Keep a separate bank account for the business from day one. This saves hours at tax time.
03Keep every receipt and invoice. Use any bookkeeping app rather than a shoebox.
04Set aside a third of what you earn for tax, in a separate pot, from the first payment.
05Consider a limited company only when profit makes it worthwhile. Sole trader is simpler and cheaper to start.
06Get insurance if you advise, handle client property or work on site.
Do not incorporate a company, register a trademark and open a business account for an idea nobody has paid for. That order costs money and proves nothing.
Ten hours a week, spent properly
The constraint is real and it is manageable if you are honest about what those hours are for. Most people spend them on the comfortable work.
—Half the time on getting customers. Outreach, conversations, following up. This is the uncomfortable half and the one that matters.
—A quarter on delivering for the customers you have.
—A quarter on everything else: admin, invoices, the website, the tools.
—Zero on logo design, brand colours and choosing between project management apps.
Where the hours come from
—Two weekday evenings, blocked and treated as appointments.
—One weekend morning, which is usually the most productive block.
—Batch similar work. Outreach in one session, delivery in another.
—Accept that it is temporary. Ten hours a week for a year, not forever.
What to build, and when
01After customer one: a one-page website saying what you do, for whom, and how to contact you. An afternoon.
02After customer three: a simple contract or terms, an invoicing tool, and a way to take payment.
03After customer five: write down how you deliver the work, so it becomes repeatable.
04After customer ten: raise your prices, and start turning down work that does not fit.
05At any point: an email list, because it costs nothing and compounds.
Notice that nothing gets built before somebody pays. Building first is how people spend a year on a business that never had a customer.
When to go full-time
The honest thresholds, in the order they matter.
—The side business consistently earns 60% to 70% of your salary, for at least three months, with the hours available to grow it.
—You have six months of personal expenses saved. Not three. Client payment cycles are slower than salary.
—Demand exceeds the hours you have, so the constraint is genuinely time rather than sales.
—You have repeat customers or a pipeline, not a run of one-off projects.
—You have tested that you can find new customers, rather than relying on the first three who found you.
Leaving because you hate your job is not a business plan. Leaving because the business is capacity-constrained is.
The failure modes I see most
—Building for three months before selling anything.
—Perfecting the brand while having no customers.
—Charging too little, which attracts the worst clients and makes the maths impossible.
—Working the ten hours on whatever feels productive rather than on sales.
—Telling nobody, so the only people who could buy from you do not know it exists.
—Quitting the job too early, then taking bad work out of necessity.
A side business is not a smaller version of a full-time business. It is a test, and the test is whether anybody pays.
Frequently Asked Questions
Do I need to tell my employer?
Check your contract. Many require disclosure or permission for outside work, and finding out after you have clients is the wrong time.
How much time does a side business need?
Eight to twelve hours a week, consistently, with about half of it spent on finding customers rather than on delivery or admin.
Should I register a company straight away?
No. Register as self-employed, keep it simple, and consider a limited company when profit makes the structure worthwhile.
How do I find my first client?
Tell twenty people who fit your target description, individually and directly, at a real price. Two yeses is a business.
When should I quit my job?
When the business consistently earns 60% to 70% of your salary, you have six months of expenses saved, and time rather than demand is the constraint.
What is the most common mistake?
Building the website, brand and company structure before selling anything, which produces three months of work and no evidence.
Before You Go
Check the contract, pick something sellable in three weeks, tell twenty people, deliver it manually, and build the machinery around whatever people paid for.
Then keep the job until the business is constrained by your time rather than by demand. Validating without spending money is the next step, and pricing is the decision most people get wrong first.
Sell in three weeks. Build afterwards.
WRITTEN BY TARIQ SALLAM
Marketing Consultant. Entrepreneur. Content Creator.
I'm a marketing consultant, entrepreneur and content creator. I help businesses grow through practical marketing, websites, SEO, content and AI.
More About Tariq →
Keep reading