How to get your first customers: the first 10 are hand to hand
Guide · Updated July 2026
Your first 10 customers will not come from a channel. They come one at a time, from four places: warm intros, the communities where your problem is already discussed, cold outbound the founder runs personally, and a few design partners on explicit terms. This page is those four sources in working order, the arithmetic behind the cold path, and the terms that keep early discounts from fooling you.
There is no channel for the first 10
Why the founder sells first, and in what order.
At 10 customers, nothing is a funnel. Ten data points cannot tell you which subject line works or which channel converts. What they can tell you, if the founder is in every conversation, is who buys, why they buy, and in what words. The first 10 are a research project that happens to produce revenue, and both halves matter.
The founder does the selling personally. The two arguments are made in full on the founder-led sales page, so here they are only in summary. The calendar math: an early sales hire spends months of ramp, at full cost, testing messages nobody has proven. The learning: a no to a founder comes with a reason, and the reason changes the product; a no to a rep rarely makes it back to the roadmap.
Warm network first, then cold. Warm intros close fastest and teach the least. Cold outbound closes slowest and teaches the most, because a stranger's yes is the only yes that proves the pitch rather than the friendship. Work them in that order, and do not stop at warm.
Where first customers actually come from
Four sources, why each converts, and where each tops out.
| Source | Why it converts | Ceiling |
|---|---|---|
| Warm intros | Borrowed trust. A vouch from someone the buyer already knows does the work a cold message spends weeks earning, so meetings come easy and close fast. | A first-degree network is finite and does not refill. Friends also buy for reasons strangers never will, so warm wins prove less than they feel like they prove. |
| Communities where the problem is discussed | You meet buyers while they are describing the problem in their own words, which hands you both a prospect and the vocabulary for every later message. | Rate-limited by community norms. Showing up daily reads as farming, and open selling gets you removed. A listening post, not a pipeline. |
| Founder-led cold outbound | The only source where you choose who to sell to, so it tests the ICP instead of your social graph. A researched founder message is a costly signal, and senior people notice it. | Founder hours. The calendar math caps hand-built quality near 20 prospects a week; it is also the one source that later scales into a channel. |
| Early design partners | They trade money for influence: a discount in exchange for shaping the roadmap. They engage deeply and forgive rough edges, which is what an unfinished product needs. | A handful at most, each one expensive in founder time. On the wrong terms, discounted revenue fakes validation instead of providing it. |
Read the ceilings as sequencing. Warm intros run out, communities are rate-limited, and design partners cap at a handful. Founder-led cold outbound is the only source you can turn up on purpose, which is why the rest of this page is mostly about doing it well.
The cold outbound path in practice
The list, the message, and what replies really run.
Build the list small and precise. For the first 10, a list of 50 companies that share a trigger beats a thousand that share an industry code. Every name should carry a reason you could say out loud: they hired for the problem, they shipped around it, they posted about it. The list is your ICP written as names, and if you cannot fill 50 rows, the ICP is not specific yet; finding that out now is the cheap version. The sending mechanics, verification and bounce hygiene through subject lines and follow-ups, are on the B2B SaaS outreach page.
Send proof of work, not claims. A first-customer message earns its reply by carrying evidence that someone spent real time on this one company: an observation from their changelog, a broken flow you actually ran, your product already working on their business. The strongest form is the last one, which we call demo-led outbound: stop claiming a demo would be worth their time and include the demo.
Expect single digits, and read every reply. At scale and average quality, measured cold email replies now run under 1%; Belkins measured 0.45% across 7.5 million emails sent in 2025. Small, heavily personalized campaigns still measure between 5% and 18% in the public studies. The full set, with methods and caveats, is on the cold outbound benchmarks page. On a 50-name list, that spread is the difference between zero conversations and a real pipeline, which is why the message work above is not optional.
Where our product fits. driftwood (driftwood.sh) is this path as a product: for each prospect, the agent does the research, builds a working demo of your product for their business, and drafts the message for you to approve. Our own first customer came exactly the way this page describes. Autosana (YC S25) is the company where driftwood's founder had been the founding engineer; the warm relationship opened the door, and what turned it into a paying customer was the agent running this cold outbound path on the same lead list Autosana's founders had been working with templates.
Reply rate before
under 1%
Reply rate, week one
over 14%
On that list, replies went from under 1% to over 14% in week one at Autosana (YC S25), including founders who had ignored more than four months of prior outreach. One company, one week, observed once; the full account is in the case study.
Design partners and discounts
What to trade for feedback, and what never to trade.
The trade, stated plainly. A design partner is an early customer who pays less and gets more say. Done well, it is the best version of a first customer: engaged, forgiving, and loud about what is missing. Done vaguely, it turns into free consulting with a logo attached. The difference is whether the honest version of the trade fits on one page and both sides can recite it.
What to trade.
- Price, for a period. A real discount with a written end date. The product has a price; the partnership has a term.
- Roadmap attention. Their problems get priority where they overlap the segment you are building for.
- Access. A direct line to the founders, and early builds before anyone else sees them.
What never to trade.
- The price of zero. A partner who pays nothing is a user interview. Payment is the signal the first 10 exist to produce; do not switch it off.
- Exclusivity. Locking yourself out of the segment you are trying to learn is the most expensive discount there is.
- The roadmap itself. Build for the segment their problems represent, not for their spec. One customer's spec makes you an agency.
- An open-ended discount. Founding-customer pricing forever reprices the product for everyone who hears about it. End dates, in writing.
Put it on one page. The price and the end date, what they owe (a standing call, real usage, a reference if it works), and what you owe (the attention and access above). None of it needs a lawyer; it needs to be written down before either side gets disappointed.
When not to scale
The first-10 lessons come before channel spend.
Spend amplifies whatever you have. A paid channel does one thing: it puts a message in front of more people. If the first 10 have not yet taught you which message, spend amplifies a guess. The lessons the first 10 exist to produce are concrete: who buys, why they buy, the words they use for the problem, what they compare you to, and what they actually pay.
Three signs you are not ready. You cannot state the ICP in one sentence that changes the first line of the message. You cannot name where each of your last five customers came from. Nobody has renewed yet, so you do not know whether the product keeps the customers a channel would bring. Any one of the three means the money is better spent on more hand-to-hand conversations.
When you are ready. The channel-by-channel comparison, what each costs and when each works, is on the B2B growth channels page. The handoff arithmetic, when the founder motion becomes a playbook someone else can run, is on the founder-led sales page.
Common questions
The questions founders ask about the first 10.
Product questions about driftwood are on the FAQ.
How do I get my first 10 customers for a B2B SaaS?
In order: warm intros from your own network and your investors', the communities where your problem is already being discussed, founder-led cold outbound on a small researched list, and one or two design partners on written terms. The founder runs all four personally. At this stage there is no channel, only conversations, and the lessons those conversations produce are as much the point as the revenue.
Should my first customers get the product for free?
No. Discount as far as you need to, never to zero. Payment, even a small amount, is the signal that separates a customer from a polite user interview, and the first 10 exist to produce exactly that signal. A free pilot with a written end date and a real price on the other side is the closest thing to an exception.
How many people do I need to contact to get my first customers?
Hundreds, not thousands. At a 5% reply rate, which the public studies say is a fair planning number for small researched campaigns, 20 sends produce one reply; assume one booked meeting per three replies, the same planning assumption as the founder-led sales funnel table, and the first 10 customers sit somewhere in the first few hundred carefully chosen prospects, depending on how well the list matches the ICP. The studies behind the assumptions are on the cold outbound benchmarks page.
Are customers from my personal network real validation?
Half. The money is real and the feedback is real, but people who know you buy partly on trust, and trust does not generalize. Warm customers prove the product can be used; strangers prove it can be sold. That is what the cold outbound path is for, and why it belongs inside the first 10 rather than after them.
When should I start scaling past the hand-to-hand stage?
When the motion is documented and repeatable: you can say who buys, why they buy, and which message gets replies, with numbers rather than a feeling. Before that, channel spend amplifies guesses. The channel-by-channel comparison is on the B2B growth channels page, and the arithmetic for handing the motion off is on the founder-led sales page.
Your first customers, hand to hand
driftwood runs the cold path on this page: research, a working demo per prospect, your approval on every send. Twenty minutes, and we will show you one built on your product.
Book a demo