Finding Your First Clients
Find clients on platforms and directly: write a message that gets replies, work out the odds of outreach on sample numbers, compare channels per hour and avoid common scams.
In this lesson you will
- List where clients come from
- Write a message that gets replies
- Work out the odds of outreach
- Compare channels per hour and avoid scams
About 23 minutes
Dev has an offer and nobody to sell it to
Dev has an offer, three samples and a profile, and no clients. This is the hardest moment for most freelancers. The good news is that finding clients is a skill with a method, not a matter of luck. It is also partly a numbers game, and the numbers are easier to bear when you know them in advance.
Where clients come from
- Platforms and job boards. Sites where clients post work and freelancers apply. They bring many options and a lot of competition. Each has its own fees and rules, which change, so read the current terms.
- Direct outreach. Writing to businesses or people you think need your service. It takes effort, and you control the message.
- Referrals. Past clients, friends and colleagues who recommend you. Usually the best clients, and the slowest to build at first.
- Being visible. Helpful posts, a small site, a profile that people find. This builds slowly.
- Local places. Shops, clubs, schools and neighbours. Often overlooked and often friendly.
A message that gets a reply
- Find a business you can actually help and look at its website or page.
- Say in one line who you are, and name one specific thing you noticed.
- Offer something concrete, such as a short sample fix, and a clear next step.
- Keep it short, polite and personal. Never send the same message to everyone.
- Follow up once after about a week, then move on kindly.
Count the odds
The program below uses made-up odds for one message: 15 per cent reply, half of those want a call and 40 per cent of the calls become clients. Multiplied, that is 3.0 per cent of messages leading to a client. With 40 messages, you would expect about 1.2 clients on average. The chance of at least one client is 70 per cent, and you need about 23 messages to reach an even chance of one. Writing 40 messages at about 20 minutes each takes 13.3 hours.
import math
# MADE-UP chances for one outreach message
reply, call, hire = 0.15, 0.50, 0.40
p = reply * call * hire
print("chance a message becomes a client:", f"{p:.1%}")
messages = 40
print("expected clients from 40 messages:", round(messages * p, 2))
print("chance of at least one client:", f"{1 - (1 - p) ** messages:.0%}")
need = math.ceil(math.log(0.5) / math.log(1 - p))
print("messages for an even chance:", need)
minutes = 20
print("hours for 40 messages:", round(messages * minutes / 60, 1))
# MADE-UP month of results: (channel, hours spent, clients won)
channels = [("job boards", 12, 1), ("direct emails", 10, 2), ("referrals", 3, 1)]
ranked = sorted(channels, key=lambda c: c[2] / c[1], reverse=True)
for name, hours, clients in ranked:
print(f"{name}: {clients / hours * 10:.1f} clients per 10 hours")
print("hours in all:", sum(c[1] for c in channels), "| clients:", sum(c[2] for c in channels))
chance a message becomes a client: 3.0%
expected clients from 40 messages: 1.2
chance of at least one client: 70%
messages for an even chance: 23
hours for 40 messages: 13.3
referrals: 3.3 clients per 10 hours
direct emails: 2.0 clients per 10 hours
job boards: 0.8 clients per 10 hours
hours in all: 25 | clients: 4
Look at what the maths says. Even with decent odds, a small batch of messages may produce nobody, and that is normal. A quiet first month is not proof that you are no good. It means the sample was small. The numbers also show why steady effort beats a one-off burst: more messages, sent over time, raise your chance in a way that no single perfect message can.
The second half compares three channels from a made-up month. Referrals won 3.3 clients per 10 hours, direct emails 2.0, and job boards just 0.8. Altogether Dev spent 25 hours and won 4 clients. The lesson is not that job boards are bad, since for some people they work well. It is that you should measure your own results per hour and spend more time where it pays. Keep a simple list of where each lead came from.
Beware of scams
- Never pay a fee to be given work.
- Be careful with a client who sends a cheque for more than the price and asks you to send back the difference.
- Keep chats and payment on the platform if it protects you, and read its current rules.
- Be wary of vague jobs, rushed requests and clients who will not name their business.
- Trust your instincts. A good client can explain who they are and what they need.
Sending ten messages, hearing nothing and quitting. Ten messages is far too small a sample. Plan a number, such as 40 over a month, and review the results after that.
The reply, call and hire rates, the hours and the channels are made up. Real rates vary enormously by service, place and message. Platforms change fees and rules often, so none are quoted. Nobody can promise that you will win clients.
Practice for this lesson
Check what you learned
Five quick questions. Pick an answer and you will see straight away whether it is right and why.
What does the 3 per cent chance per message mean?
- You will win a client every 3 messages
- About 3 in 100 messages lead to a client
- Messages never work
- Three clients per message
The three rates multiply to 3 per cent in the example.
Why is a quiet first month not proof of failure?
- Messages are banned
- Clients never reply
- The numbers are fixed
- The sample is small, so zero is quite possible
With low odds and few messages, no result is common.
Which channel gave the most clients per hour in the example?
- Job boards
- Referrals
- Direct emails
- None
Referrals were fastest per hour in the sample.
What should you never do?
- Follow up once
- Keep a lead list
- Read the rules
- Pay a fee to be given work
Pay-to-work offers are a common scam.
Why measure results per hour?
- To spend time where it pays
- To avoid tax
- To look busy
- To raise your price
It shows which channel to favour.
Summary
- Clients come from platforms, outreach, referrals, visibility and local places.
- Outreach is a numbers game, so plan a number and review.
- Measure results per hour for each channel.
- Never pay to get work, and watch for overpayment scams.
Questions
Should I use freelance platforms?
They are one option and suit some people well. Read each platform's current fees and rules, and do not rely on one source of clients.
How many messages should I send?
Pick a number you can keep up, such as a few a day, and review after a month. Small samples are noisy.
Are these rates typical?
No. They are made-up numbers to show the method. Your own rates will differ, so measure them.
