How to Get Your First Customers With No Audience

You do not need an audience. You need fifty specific people and something worth saying to them.

That is the honest answer, and most people resist it, because building an audience feels productive and safe while sending direct messages to strangers feels exposed. But an audience is a slow, indirect way to reach the same person you could contact today. Posting for six months to build a following of 800 people, of whom maybe 12 are buyers, is a worse use of the same six months than talking to 300 buyers directly.

This article covers the whole method: how to narrow the target, what offer to build, what proof you need before anyone will take you seriously, how to find and contact 50 people, what to say, what the numbers usually look like, and what to do when nobody replies.

The short answer

Get your first customers with no audience by doing five things in order:

1. Pick one narrow customer. Industry, location, problem and ability to pay. Not "small businesses" — "dental practices in Manchester with two or more locations and no one running their recall reminders."
2. Build one offer. One deliverable, one price, one turnaround time. Written in a single sentence.
3. Make one piece of proof. A sample, a mockup, a before/after, a two-page mini-audit. Something that shows the work instead of describing it.
4. List 50 real people or businesses that match the target, with names and contact details.
5. Send 50 personalised messages that open with something you noticed about their business, not your background. Then ask every conversation — including the ones that say no — for one introduction.

Then repeat the one channel that produced conversations before adding a second one. That is the whole thing. The rest of this article is why each step works and how it goes wrong.

Why "no audience" is not actually your problem

When you say "nobody knows my business", you are describing a distribution problem. But the thing blocking your first sale is almost never distribution. It is usually one of these:

Audience solves exactly one of these — the first — and it solves it slowly and expensively. Direct outreach solves it in a week. So start there, and build the audience later with the material your first customers give you.

Step 1: Narrow until it feels uncomfortably small

The single biggest reason first-customer outreach fails is that the sender does not know precisely who they are writing to, so the message has to work for everyone and therefore lands on no one.

Narrow along four axes. All four, not two.

• Industry or type. Not "professional services." Independent physiotherapy clinics. Specialty coffee roasters. Regional freight brokers.
• Location or context. A city, a region, a platform, an event, a trade body. Something that lets you find them in a list.
• A specific problem you can name. Not "growth." "You take bookings by phone and lose the ones who call after 6pm."
• Ability and authority to pay. Can this business plausibly write a $1,500 check without a committee, and does the person you are contacting have the authority to do it?

The test: can you sit down right now and name 20 real businesses that fit? If you cannot, you have not narrowed — you have described a demographic. Narrow again.

Narrow feels like it shrinks the market. It does the opposite. A message that says "I noticed your booking page still asks people to email you for availability" gets read. A message that says "I help businesses grow online" gets deleted. And you can change the niche next month; you are not marrying it. If you are still unsure whether the underlying idea holds up at all, do the cheaper work first and validate the business idea before you spend six weeks on outreach.

What "narrow" looks like in practice

A hypothetical bookkeeper who has been trying to reach "small businesses" narrows to: construction subcontractors in one county, 3–15 staff, who invoice through spreadsheets and file VAT late. Suddenly the message writes itself, the list is findable through trade directories and the local authority's contractor registers, and the problem — late VAT, penalties, cash flow guesswork — has a number attached to it.

Step 2: One offer, one price, one turnaround

Your first offer should be almost insultingly specific. Not a menu. Not "packages from." One thing.

A good first offer has four parts:

Price it so that one sale is worth the effort of 50 messages. For most service businesses starting out, that means a first offer somewhere between $500 and $3,000 (or the local equivalent) — big enough to matter, small enough to be approved by one person without a procurement process. Below about $300, the sales effort costs more than the sale. Above about $5,000, first-time buyers who have never heard of you usually want references you do not have yet.

The exact right number depends on your market and what your competitors charge. Go and look — ask three people in your target market what they last paid for something similar, and check what published-price competitors list.

Under-pricing is the more common mistake, and it is expensive twice: you earn less, and cheap prices signal inexperience to exactly the buyers who could afford you. If you are not sure whether you are building the right kind of offer at all — one-off project, retainer, productised service — work through choosing a business model before you commit to a price you will have to defend.

Step 3: Build proof before you have customers

Nobody buys from a business with no evidence. But evidence does not have to mean customers. It means something concrete they can look at.

Pick one and build it this week:

1. A sample. Do the work for a business you are not contacting yet. A sample email sequence, a sample landing page, a sample month of bookkeeping cleaned up from mock data.
2. A mockup. Redesign one page of a real prospect's site and send it to them. This is the highest-converting cold outreach asset in existence and almost nobody does it, because it takes 40 minutes.
3. A before/after. From your own business, a friend's business, or a volunteer project. Two screenshots and one number.
4. A free mini-audit. A repeatable 2-page document: five things you found, what each is costing them, what fixing it involves. Keep it under an hour of work per audit.
5. A live demo. For anything software-shaped, a working thing beats a deck. If you are still deciding what to build, get to a minimum viable product rather than a full build.

Two rules. Time-box it. The proof asset should cost you a day, not a fortnight — it exists to start conversations, not to be your portfolio. And do not work for free at full scale. Free samples that take an hour are marketing. Free three-week projects attract people who were never going to pay and teach you nothing about whether your price works.

Step 4: Build the list of 50

Fifty is the number because it is large enough to produce a signal and small enough to finish in two days. Under 20, you cannot tell the difference between a bad offer and bad luck. Over 100, personalisation collapses and you start sending templates, which is where cold outreach earns its reputation.

Where to find 50 real names, in rough order of quality:

Track it in a spreadsheet with six columns: name, business, contact, the specific thing you noticed, date contacted, outcome. That last column is the whole point — after 50 rows you can see which segment replied and which did not.

Two practical notes. First, check the rules where you are: unsolicited B2B email is broadly permitted in many places but regulated (GDPR and PECR in the UK and EU, CAN-SPAM in the US), and unsolicited messages to consumers are treated more strictly. Look up your own jurisdiction's current position, or ask a solicitor, before you send at volume. Second, tools like Apollo, Hunter and Clay can find email addresses at scale — they are useful for finding the address, and dangerous if you let them write the message.

Step 5: Write messages that lead with their problem

Here is the structure that works. Four sentences, under 120 words.

1. The specific observation. One thing you noticed about their business. Not their industry, theirs.
2. The cost of it. What that thing is probably costing them, in money, time or customers. Reasoned, not invented.
3. The offer. One deliverable, one price, one turnaround.
4. A small ask. Not "book a call." Something with almost no cost — "want me to send the two-page version?"

What kills the message:

• Opening with yourself. "I'm a freelance developer with 8 years of experience" is the most common first line in cold outreach and the reason most of it fails. Nobody is reading a message from a stranger to learn about the stranger.
• Asking for 30 minutes. Thirty minutes is expensive. Two minutes of reading is cheap.
• Attachments and links in the first message. They trip spam filters and they ask for effort before you have earned it.
• Fake familiarity. "Hope you're having an amazing week" reads as a template because it is one.
• No specific ask. "Let me know if you're interested" puts the work on them.

Follow up twice, four to seven days apart, adding something each time rather than nudging. A second message that contains the mockup you mentioned works. A second message that says "just bumping this" does not. Then stop. Two follow-ups is persistence; five is a complaint.

Personalise the first two sentences properly and template the rest. Ten researched messages beat 200 generic ones — but do not confuse personalisation with padding. One accurate specific detail is enough.

Where AI helps and where it ruins this

AI is genuinely good at the research half: summarising a prospect's site, pulling the specifics out of their reviews, cleaning your list, drafting the structural skeleton of the message. It is bad at the part that makes outreach work, because the value of "I noticed your booking page loses after-hours calls" comes from the fact that a human actually looked. Generated personalisation reads as generated, and prospects have got very good at spotting it.

Use it to go faster on the boring 80%. Write the first two sentences yourself. If you want the broader picture of where this fits, see how to use AI in a small business.

What the numbers usually look like

There is no reliable public figure for cold outreach conversion — it varies enormously by industry, price point and how well you have narrowed. What follows is the working range we use for planning, not a promise. Treat it as a way to notice when something is broken.

The important diagnostic: negative replies are information, not failure. Fifty messages producing zero replies means the targeting or the opening line is wrong. Fifty messages producing eight "not right now" replies means the targeting is right and the timing or offer needs work — a much better position.

If the answer is "not right now," ask what would have to be true for it to be right, and ask when to come back. Put the date in your calendar. A "no" with a date attached converts more often than a cold new name.

Ask for one introduction, every single time

This is the step almost everybody skips, and it is the one that compounds.

At the end of every conversation — sold, not sold, ghosted-then-replied — ask one question: "Who else do you know with this problem?" Not "do you know anyone", which invites "not off the top of my head." Ask for one specific person, and make it easy: "Is there one other clinic owner in the area you'd point me to?"

Roughly one in five people will give you a name if you ask directly. Those names convert far better than cold ones, because they arrive with a borrowed reputation. Fifty cold contacts that produce ten introductions is a warm list you did not have last month, and it is the beginning of the audience you were worried about not having.

When someone gives you a name, ask them to forward a message you write for them. Never ask a busy person to compose an introduction from scratch.

Turn the first win into the next three

The first customer is worth more than the money. Extract all of it, within two weeks of delivering, while they are still pleased:

1. Ask for a testimonial with a number in it. "It's great" is worthless. "Cut our quote turnaround from four days to one" is a sales asset. Offer to draft it for them from things they already said, and let them edit — this triples the response rate and is entirely honest as long as they approve the final wording.
2. Write a one-page case study. Situation, what you did, what changed, over what period. Even one is enough to change how the next 50 messages read.
3. Ask for two introductions, not one, now that you have delivered.
4. Offer the next thing. The cheapest customer to win is the one you already have. Have a second offer ready before you finish the first project.
5. Write down how you did it. The steps, the wording, the delivery checklist. This becomes your process, and eventually the thing you hand to someone else — see how to write SOPs when you get there.

This is the moment when the method compounds. Every customer after the first should cost you fewer messages than the one before.

If you want the wording for all of this — the outreach templates, the testimonial request that actually gets answered, the list-building checklist and the decision tree for what to change when 50 messages produce nothing — that is what The First 10 Customers Guide is. It is the fill-in version of this article: scripts you can send today, the follow-up sequence with timings, and the recovery paths for the four common ways this stalls.

What this costs

Direct outreach is the cheapest customer acquisition available to a new business, which is the main argument for doing it before anything else.

Compare that with the alternatives at the same stage. Paid ads typically need $1,000–$3,000 of spend before the data means anything, and they perform badly when you cannot yet describe your offer precisely. Content and SEO usually take six to twelve months to produce leads. Both are reasonable second moves. Neither is a reasonable first move when you have no customers and no clarity, because both require you to already know what to say — and the way you find that out is by talking to 50 people. For the wider picture of setup costs, see what it costs to start a business.

The five ways this goes wrong

You send 50 messages to 50 different kinds of business. Then no result tells you anything. Keep the segment constant across the whole batch. Change one variable at a time.

You wait until everything is ready. The website, the logo, the LLC, the pricing page. None of those produce a customer. A one-paragraph offer and a working email address are sufficient. Get the rest in order with the business launch checklist in parallel, not first.

You go quiet after the first no. The first ten rejections in a batch of 50 are statistically normal. Stopping at message 12 is the most common cause of "outreach doesn't work for my business."

You add channels instead of repeating the one that worked. If email produced two conversations and Instagram produced none, the answer is more email, not TikTok. Repeat the working channel until it visibly stops working. Adding channels feels like progress and is usually avoidance.

You discount to close the first deal. A 50% discount to win customer one sets your anchor with the person most likely to refer you. Instead of cutting the price, cut the scope: same rate, smaller first project.

What to do this week

A realistic five-day version:

That is roughly 20 hours of work spread across two weeks. Most people spend more than that on a logo. If you want the longer arc of what happens before and after this fortnight, from idea to first customer covers the surrounding sequence.

When outreach is the wrong first move

Be fair about the exceptions. Direct outreach is the default, not a law.

• Low-price consumer products. A $25 product cannot support a 12-minute personalised message. Use marketplaces, local retail, markets or paid social with small test budgets.
• Genuinely local walk-in businesses. A café gets its first customers with signage, Google Business Profile and being open. Outreach applies to wholesale accounts, not to footfall.
• Regulated buying. If your buyer must run a tender, outreach starts the relationship but the procurement process sets the timeline. Expect months, and qualify hard for budget.
• You already have a relevant audience. If you have 3,000 followers in the exact niche, ask them first. Outreach is the substitute for that, not a superior alternative.

Even in these cases, the underlying discipline holds: one narrow customer, one clear offer, one piece of proof, and a countable number of direct conversations.

The short version

• You do not need an audience. You need 50 specific people, a narrow definition of who they are, and one thing you noticed about each of them.
• One offer, one price, one turnaround. Somewhere in the $500–$3,000 range for most service businesses, because that is what one person can approve alone.
• Build one piece of proof in a day — a sample, a mockup, a mini-audit. Show the work rather than describing it.
• Expect roughly 5–15 replies and 0–2 sales per 50 researched messages. Zero replies means the target or opening line is wrong; several "not right now" replies means you are close.
• Ask every conversation for one introduction, and turn the first win into a testimonial with a number in it, a one-page case study and two more introductions.
• Repeat the channel that produced conversations before adding another one. Adding channels feels like progress and usually is not.

The First 10 Customers Guide is the $10 companion to this: the outreach and follow-up scripts, the 50-contact list template, the qualifying questions, the testimonial and referral wording, and a decision tree for what to change when a batch of 50 produces nothing.