Starting a service business: the first 90 days, in the order that actually works

Days 8 to 14: define one service, for one customer

New businesses lose months trying to be available for everything. Pick one service you can deliver well, for one identifiable customer, in one defined area. "Residential gutter cleaning and repair for homes in these six ZIP codes" beats "general exterior services, anywhere". A narrow offer is easier to price, easier to explain, easier to route, easier to advertise and far easier to get referred, because people repeat simple descriptions and forget complicated ones. You can widen later, from a position of knowing what sells.

Days 8 to 14: price from cost, not from the competition

Copying the price of the established company down the road is how new businesses go broke while staying busy. Build the number from your own cost of an hour: take your target annual income, add every fixed cost for the year — insurance, vehicle, phone, software, advertising, tools, taxes — and divide by the billable hours you can realistically sell. Most owners are shocked here, because realistic billable hours are closer to 1,100 a year than 2,000, once driving, estimating, invoicing and dead time are removed. That number is the floor. Anything you charge below it is a job you pay to perform.

Days 15 to 30: the first ten customers

The first ten do not come from advertising. They come from a list you write by hand: everyone who knows you, every former colleague, every neighbor, every local business owner you deal with, every trade contact adjacent to your work. Tell them plainly what you now do, where you work and what a typical job costs. Ask directly whether they know anyone who needs it. This is uncomfortable for exactly one week and it produces work faster than any channel you can buy.

The goal of these ten is not profit. It is evidence: photographs, reviews, references and an honest measurement of how long jobs actually take compared to what you assumed.

Days 15 to 30: build the proof while you work

From the very first job, photograph before and after, ask for a review the same day, and ask the customer to name the service and the neighborhood in it. Ten specific reviews from your first month are worth more than any marketing you will do in year one, and they are free. Six months later, when you are competing against companies with a hundred reviews, the thing that separates you is not quantity but recency and specificity.

Days 20 to 30: be findable

Before a website, claim and complete the business profile: categories, service area, hours, services with descriptions, payment methods, photos. Completing it well outperforms a website in the first months, because it is where the search happens. A single-page website with your service, area, phone number, license and insurance status and ten photos is enough for the first quarter. Anything more elaborate is a decision better made once you know which service is selling.

Days 31 to 60: answer faster than everyone else

Response time is the only competitive advantage available to a new business on day one, and it is the one most established competitors have given up. Answer live during working hours. Return every missed call with a text within five minutes. Send the quote the same day, not "by the end of the week". In most residential trades, the company that responds first wins a disproportionate share of the jobs, regardless of price or experience.

Days 31 to 60: the quote that gets accepted

A quote that wins has six parts: what will be done in plain language, what is excluded, the price, the timeline, the payment terms and how long the price holds. Send it as a document, not as a text message paragraph. Include one photograph of similar completed work. Follow up once after two days and once after a week, then stop. New owners either never follow up or follow up until they are irritating; two touches is the balance that closes without damage.

Days 31 to 60: get paid on time from the start

Set payment terms before the first job, not after the first slow payer. For residential work, payment on completion is standard and there is no reason to offer anything else. For jobs above a few thousand dollars, take a deposit for materials. Accept cards even though the fee stings, because the alternative is chasing checks. Invoice the day the job ends — invoices sent a week later get paid three weeks later. Cash flow, not profit, is what kills first-year service businesses.

Set the tax habit in month one

The first year ends badly for a surprising number of otherwise successful service businesses, and the reason is almost always the same: nothing was set aside. Someone who has only ever been an employee is used to tax arriving already deducted. Working for yourself, the full amount is yours to hold and yours to send, and it includes self-employment tax on top of income tax. The habit that prevents the problem is mechanical rather than clever: open a second business account, move 25 to 30 percent of every payment into it the day the money lands, and never treat that account as available. Pay quarterly estimates rather than waiting for April, which avoids penalties and, more importantly, avoids the situation where the money owed has already been spent on a truck payment.

The same discipline applies to bookkeeping. You do not need an accountant in month one, but you do need every expense captured, because the deductions that reduce a service business's tax bill — mileage, tools, insurance, phone, a home office, materials — are precisely the ones that go unclaimed when receipts live in a glove compartment. A weekly fifteen-minute habit of photographing receipts and categorizing the bank feed costs almost nothing and routinely saves thousands at filing time. It also means that when you eventually want a vehicle loan, a line of credit or a buyer for the business, the records that prove what you earn already exist.

Days 31 to 60: track four numbers, not forty

Every week, record four things: quotes sent, quotes accepted, revenue collected and hours worked. From those you get your close rate, your average ticket and your real hourly rate. Four numbers on one sheet will tell you more in a month than any software you could buy, and they will surface the two problems that are usually invisible — quoting too low and closing too little.

Days 61 to 90: turn jobs into repeat work

A first-year business that treats every job as a one-off has to win every customer twice. Before leaving any job, book or schedule the next contact: the seasonal return, the maintenance interval, the second phase the customer mentioned. Even an informal "I'll message you in October when it's time" converts at a rate advertising cannot match, and it costs one sentence.

Days 61 to 90: add exactly one paid channel

Only now, with a close rate and an average ticket known, does paid acquisition make sense — because only now can you tell whether a lead that costs $60 is cheap or ruinous. Start with one channel, run it for sixty days, and judge it on booked jobs, not clicks or calls. Running three channels simultaneously in month three guarantees you will not know which one worked.

Days 61 to 90: the first hire question

Most owners hire too early, out of exhaustion, and regret it within a month. The test is objective: are you turning away work you could sell, consistently, for four weeks or more? If yes, the math is straightforward — a helper who costs $3,000 a month must enable enough additional jobs to cover that plus materials and still leave margin. If no, the right move is raising prices, which produces the same relief without the payroll risk.

Learning to decline work

New owners accept everything, and it is the most common reason the first year feels like a treadmill. Decline jobs far outside your service area, jobs outside your defined service, customers who negotiate before you have finished describing the work, and anything requiring a license you do not hold. Each of those declined jobs returns hours you can sell to the customers you actually want.

What a realistic first ninety days looks like

A solo residential operator with a $95 hourly target, working the sequence above, typically ends month one with 6 to 10 jobs and $2,000 to $4,000 collected, mostly from personal network. Month two, with reviews accumulating and response speed working, 14 to 20 jobs and $5,000 to $8,000. Month three, with repeat work and one paid channel, 20 to 28 jobs and $7,000 to $11,000. That is the shape: slow, then compounding. Anyone promising a full calendar in week two is selling something.

The five mistakes that cost the most

Buying equipment before having customers. Pricing against a competitor whose costs you cannot see. Mixing business and personal money. Working without insurance "until things get going", which is precisely when a claim would end the business. And spending month one building a website while zero people know the company exists. All five are avoidable, and all five are common.

Knowing whether it is working

At day ninety, three signals tell you the truth. Your close rate on quotes is at or above roughly one in three. At least a quarter of new inquiries come from referral or repeat customers rather than from you chasing. And your collected revenue divided by hours worked is approaching the hourly number you calculated in week two. If all three are present, the business is working and the job now is volume. If none are, the problem is almost always the offer or the price, not effort.

Where to start tomorrow

Do the four legal items this week. Write the one-service, one-customer, one-area sentence and say it out loud until it is short. Calculate your hourly floor from your own costs. Then write the list of eighty people you know and start telling them. Ninety days from now you will have the only thing that makes every later decision easy: real customers, real numbers and evidence of what sells.

Frequently asked questions

What should I do first when starting a service business?

Four legal items in week one, and nothing else: register the entity and get the federal tax ID, open a business bank account separate from personal on the same day, buy general liability insurance because commercial customers and many homeowners ask for the certificate before you start, and confirm the license requirement for your trade, state and city. Everything people describe as setup — logo, uniforms, vehicle lettering, brochures, an elaborate website — can wait sixty days without costing you a single job. Spending week one on branding while no one knows the company exists is the most common and most expensive misordering in a new service business.

How do I get my first customers without advertising?

From a list you write by hand: everyone who knows you, former colleagues, neighbors, local business owners you deal with, and trade contacts adjacent to your work. Tell each one plainly what you now do, where you work and what a typical job costs, then ask directly whether they know anyone who needs it. It is uncomfortable for about a week and it produces work faster than any channel you can buy. The purpose of those first ten jobs is not profit — it is evidence: before and after photographs, specific reviews naming the service and the neighborhood, references, and an honest measurement of how long jobs really take versus what you assumed.

How should I price when I am just starting?

From your own cost, never by copying the established company down the road, whose cost structure you cannot see. Take your target annual income, add every fixed cost for the year including insurance, vehicle, phone, software, advertising, tools and taxes, then divide by the billable hours you can realistically sell. That last number surprises most owners: realistic billable hours are closer to 1,100 a year than 2,000 once driving, estimating, invoicing and dead time are removed. The result is your floor. Any job priced below it is work you are paying to perform, and no amount of volume fixes that.

When should I hire my first employee?

The test is objective rather than emotional: have you been turning away work you could have sold, consistently, for four weeks or more? If yes, run the math — a helper costing roughly $3,000 a month must enable enough additional jobs to cover that plus materials and still leave margin, which means the calendar has to be full before the hire, not after. If you are exhausted but not turning away work, the correct move is raising prices, which produces the same relief without payroll risk. Most first hires happen too early, out of tiredness, and are regretted within a month.