The first ten customers of a company will fit on an index card, and that is the fact most first-ten advice manages to ignore. Ten does not require a channel, a launch day, or a growth strategy. It requires a couple hundred named human beings and a loop you run against them until some of them pay you.
Here is the loop, the credit arithmetic behind a month of running it, and what the whole thing costs.
Count the market before you try to sell into it
The first move is arithmetic. Describe your buyer tightly enough that a search over company and people data can return them: role, industry, headcount band, geography. Then look at how many come back.
In Rocketship the search, the count, and a preview of twenty-five of those people by name are free, with no card. That count is worth more than it looks, because it is a strategy decision wearing the costume of a data question. A description that returns twelve thousand people has not really been written yet; it describes a category rather than a buyer, and your first ten will come out of a slice of it you have not defined. A description returning forty means every deal is a relationship and outbound volume is the wrong instrument entirely. Somewhere in between you have a list you can work through deliberately, which is what the rest of this is about.
Read the twenty-five names before you spend anything. If your honest reaction is that these are not your customers, you learned it for nothing.
Four moves, and launch day is not among them
Find a named person. Write to them. Answer what comes back within a day. Get on the phone. Nearly everything else founders do in month one, the Product Hunt post, the launch thread, the waitlist page nobody signs up for, is elaborate avoidance of the first move, because the first move looks a lot like cold calling.
The loop beats the launch because it runs at a known cost and produces a rate. A launch hands you a spike you cannot reproduce and a pile of email addresses belonging to people who follow launches. The loop tells you that writing to sixty people yields some number of replies and some number of calls, and once you know that number you know what ten will take.
You do not know your rates yet. Write down a guess, run sixty, correct the guess. Guess low. Founders who budget their morale on a third of people answering are usually gone before the second sixty.
What 100 credits buys in a month
Specifics, since this is where this kind of advice normally goes soft. Launch includes 100 lead credits a month. Adding a person from the lead database into your CRM costs 1 credit. A verified phone number and email address for that person cost 5, and the 5 is charged only when they are actually found.
So the month has a shape. Add sixty people at a credit each. Read them properly. Pick the eight you would give up a Tuesday afternoon to talk to, and pull verified contact details for those eight at five apiece. Sixty plus forty is your hundred. You finish with sixty people to write to and eight you can call by name at a number that rings.
Do that for three months and you have worked through 180 companies and made roughly two dozen calls to people you chose on purpose. If ten customers are not somewhere inside that, the offer is what needs fixing, and you found it out for a few hundred dollars instead of a year.
Week three is where this usually falls apart
Sending is the easy half. Anyone can send sixty emails. The attempt dies in the third week, when a chunk of them have bounced, six people have replied, two of those replies are some version of who is this, one says ask me in March, and you are three days into a bug and have answered none of them.
The AI worker on Launch reads what comes back and answers it. Replies get classified, so an interested one is not sitting in the same undifferentiated inbox as an out-of-office. Bounces are suppressed rather than retried until the domain is cooked. Your sending domain gets checked before you burn it, which founders learn to care about immediately after burning one.
One worker, on the $24.99 plan, doing the follow-up that people with a product to ship reliably fail to do by hand.
Your phone rings twice while you are already on a call
Both callers had read your emails. Neither one leaves a voicemail, because hardly anybody does now.
Launch includes an inbound line answered around the clock. It picks up, it handles after hours, it can hand the conversation to a human mid-call when the caller deserves you instead of it, and every call comes back as a recording and a transcript you can actually read. For the first ten, those transcripts are worth about as much as the meetings, because they are the only unedited record of a stranger explaining your product back to you.
Frontier at $79 adds the scheduler, so an interested reply becomes a meeting on your calendar without a link, a back-and-forth, and a time zone mistake. Two workers, 250 credits. Mission Control at $149 runs three workers, 500 credits and unlimited apps, which is the setup for someone running this loop on behalf of clients.
The bill, and the bill you would otherwise be paying
Building and publishing an app is $0 and takes no card. The builder on its own is $12. The acquisition side, the part that finds buyers, writes to them, handles what comes back and answers your phone, starts at $24.99 a month. The price is published, you pay by card, you are live in minutes, and you cancel from the dashboard yourself. Nobody makes you sit through a demo call to hear the number.
Now price the assembled version. Lead data and enrichment from Apollo or Clay or Lusha. A sending tool such as Instantly. A CRM, and HubSpot's free tier stays free right up until the month it becomes the largest line on your card. A scheduling link from Calendly. An answering service along the lines of Ruby or Smith.ai, billed by the minute or the call.
Five subscriptions is the smaller half of that problem. The expensive half is the four boundaries between them. An enriched lead does not walk itself into the CRM. A reply does not create the task. A call transcript does not attach itself to the deal record. Somebody carries the work across every seam by hand, and at ten customers that somebody is you, when your hours are the only thing the company has to spend.
Do the first pass before Friday
Write the buyer description. Run the count. Read the twenty-five free names. If they look right, add sixty and enrich eight. Send this week, answer everything inside a day, and keep every call on record.
Ten customers will not arrive as a milestone. It arrives as a spreadsheet with ten rows, each one traceable back to a particular Tuesday and a particular email you sent. What makes that worth having is that you can run it again next month and know roughly what it will produce.