How to Get Your First 10 Customers Before You Have a Product

Ten paying customers is the smallest number that proves anything. One is luck. Three is your network. Ten strangers who gave you money is a signal — the first real evidence that the problem is worth money to someone who does not care about your feelings.

The counterintuitive part: you do not need a finished product to get there. You need a credible promise, a way to be found, and a reason to say yes now.

Why "build it and they will come" inverts the order

The default sequence is build for six months, launch, then look for customers. It fails for a structural reason: by the time you start selling, every product decision is already locked, and the first thing customers tell you is that you locked them wrong.

Selling first inverts this. Every objection you hear before you build is a feature specification handed to you for free. Every refusal tells you something about the segment, the price, or the pitch. And the ten people who say yes become your design partners rather than your first churn cohort.

The trade is that you have to sell something that does not exist yet, which feels uncomfortable and is completely normal.

Step 1 — Put up a page that makes a specific promise

Not a coming-soon splash. A real page that states who it is for, what changes for them, and what it costs.

The specificity is the whole point. "Scheduling software" converts nobody. "Stop re-confirming next week's walks over text every Sunday — $19/month" converts the exact person you want, and repels everyone else, which is equally valuable at this stage.

Your page needs four things and nothing else: the problem in the customer's words, the mechanism in one sentence, the price, and a single action. Add a testimonial slot later, once you have one that is real.

The metric that matters: visitor-to-signup rate. Under 2% and your promise is not landing. Over 10% from a targeted audience and you have something. Traffic volume is almost irrelevant at this stage — 100 well-targeted visitors tells you more than 10,000 random ones.

Step 2 — Build a list of 100 named humans

Not a segment. Not a persona. One hundred actual people with names, and a reason each one is on the list.

Where they come from depends on the business: LinkedIn search by title and company size, industry directories, subreddit and forum participants, conference attendee lists, people leaving reviews on the incumbent tool, local business listings.

The last one is the most underused. People who left a two-star review of your competitor last month have declared both the problem and their dissatisfaction in public. That is the warmest cold list that exists.

One hundred is the right number. At a realistic 10–20% reply rate and a 30–50% close rate on replies, one hundred contacts produces roughly your first ten. Fewer than that and you are reading noise.

Step 3 — Send outreach that does not read like outreach

The entire game is the first line. If it could have been sent to anyone, it will be deleted by everyone.

A cold email that works has four parts and stays under 120 words:

• A first line that proves you looked. Something specific to them — a review they left, a post they wrote, how their business is structured. Not "I loved your website."
• The problem, in their language. One sentence, and it should make them nod before they have decided whether they like you.
• What you are doing about it. One sentence. No feature list.
• A small ask. Not a demo, not a call. "Is Sunday scheduling actually the worst part, or am I aimed at the wrong thing?" People answer questions far more readily than they accept meetings.

Follow up twice, four and eleven days later, each time adding something rather than repeating yourself. Most replies arrive on the second or third touch, which is exactly where most founders have already given up.

Step 4 — Ask for money before the product is ready

This is the step people skip, and skipping it invalidates everything before it.

Enthusiasm is not evidence. "This sounds great, send me the link when it launches" is what people say instead of no. The only reliable signal is a transaction — and there are several ways to get one before the product exists:

• Pre-sale at a discount. Lifetime or first-year pricing in exchange for paying now and tolerating rough edges.
• Paid pilot. A small fee for early access plus direct influence over what gets built.
• Deposit. Refundable, small, and still enormously more informative than a mailing list signup.
• Concierge. Deliver the outcome manually for the first ten customers. They do not care that a human is doing it; they care that the Sunday problem went away.

Concierge is the most underrated of the four. It gets you paying customers immediately, and doing the work by hand teaches you exactly which parts are worth automating — which is a far better product spec than any amount of planning.

What the first ten actually tell you

Pay attention to how they bought, not just that they did.

If all ten came from one channel, you have found your channel. If they came from ten different places, you have not found it yet and should not scale spend. If nine were easy and one took six weeks of negotiation, that one is probably the wrong segment. If several asked for the same missing thing before saying yes, that is your next build, and it was specified by people holding money.

Also watch what they do after paying. Ten customers who paid and never logged in is a worse outcome than five who paid and use it daily. Revenue proves the pitch works; usage proves the product does.

Frequently asked questions

Is it dishonest to sell something that does not exist?

Only if you hide it. Say plainly that it is early, that they are getting a discount for the inconvenience, and what the timeline is. Founders who are upfront about this close more often, not less — early adopters are buying influence as much as software.

What if nobody replies to my outreach?

Almost always the list or the first line, not the product. Zero replies from a hundred contacts means the segment is wrong or your opener is generic. Change one variable at a time and send another fifty.

How long should this take?

Two to six weeks for most B2B and prosumer ideas. If you are three months in without a single payment, the problem is upstream — usually the promise on the page, which means it is time to revisit the 5-minute validation framework.

Do I need a company and a payment processor first?

No. Take the first few payments however is legal and simple where you are. Formalising the entity before you know the business exists is a way of feeling productive without being productive.

What if my product is consumer, not B2B?

The sequence holds but the numbers change: you need a larger list and the price point is lower, so ten paying users is a weaker signal. For consumer, look for ten users who came back unprompted in week two.

Run the whole sequence in one place

Page, list, outreach, pipeline — that is four tools and a spreadsheet for most founders, and the gaps between them are where the momentum dies.

Forze ships a live landing page with lead capture on your own subdomain, scouts leads that match your venture, runs the cold-email campaigns, and catches every reply in a CRM that tracks who is close to saying yes. It builds all of it from your idea brief, so the promise on the page and the first line of the email describe the same business. Free tier, no credit card.

Ship your page and start outreach free →

Not sure the idea holds up yet? Compare building an AI MVP against hiring a developer first.