How to get your first 10 customers for a B2B SaaS without an audience or network

The short version. Your first 10 B2B SaaS customers come from one painful use case, a hand-built account list, founder-led selling, and paid pilots before any audience compounds. Start narrower. If you lack a network, borrow signal from job posts, category reviews, funding notices, product launches, support forums, and regulatory deadlines, then turn those signals into specific conversations with buyers who already feel the wound.

Before you start:

• Your first 10 customers come from a narrow wound, hand-built account list, founder-led conversations, and paid pilots.
• An audience helps later; early traction comes from buyer specificity and visible pain.
• Cold outbound can work when each note names a public trigger and one costly job.
• Charge for pilots once the buyer has urgent pain, an owner, budget pressure, and a deadline.

Step 1: pick one narrow beachhead until the list feels obvious

Start painfully narrow. A founder with no audience wins by choosing a buyer group so specific that job titles, software stack, trigger event, and budget owner can be named from public evidence. I would anchor the first pass in a working ICP, then pressure test it against Your first ICP: how to pick who to sell to before you have data. The market should feel smaller than your ambition.

Your beachhead can start in G2 category pages, LinkedIn Sales Navigator filters, hiring posts, partner directories, or Slack archives where buyers complain in plain language. Use names. If the group needs a committee you cannot map, a data source you cannot reach, a compliance trigger you cannot verify, or a budget story you cannot explain, the segment is still fog, even if the product is strong.

• One buyer title
• One urgent workflow
• One public trigger
• One budget owner
• One before-state

Step 2: ask what broke this week until the same wound repeats?

Do discovery before demo. Without a network, your advantage is that no one expects polished category doctrine from you in the call, so the buyer can teach you the language competitors missed. Ask about the last broken workflow, the manual workaround, the person blamed, the tool already tried, and the moment the pain becomes boardroom-visible. Record nouns over adjectives.

The first calls are research conversations with a commercial spine. I like a paid-problem frame because it keeps courtesy meetings from poisoning the sample and keeps the founder from mistaking kindness for demand. If you launched into silence, the diagnosis in You launched and nobody came: a triage plan for zero users pairs well with this step. Done means the same wound appears without prompting.

• Last painful event
• Current workaround
• Internal owner
• Existing tool
• Costly deadline

Step 3: build a named account list until every record has a trigger

Build the list by hand. A hand-built list wins because each row carries a live clue, a buyer name, a likely metric, a competing tool, and a sentence you can defend without pretending to know their company. Pull from G2 categories, LinkedIn account search, job posts, product release notes, marketplace reviews, and conference sponsor pages. Each account needs a reason that would survive being forwarded internally.

Separate accounts from people. The account hypothesis says why the company hurts, while the contact hypothesis says why this person might own the fix, influence the owner, defend the spend, or feel the failed workaround every week. For US buyers, I would also tune the promise with Positioning an India-built product for American buyers. The list is ready when weak accounts feel embarrassing.

• Company trigger
• Buyer name
• Likely pain owner
• Current system
• Personal opening line

Step 4: send diagnosis-first outbound until replies mention the problem

Write like a diagnostician. Cold outbound still works when the email proves you saw a specific business condition before asking for calendar time, and the sequencing choices in Does cold outbound still work for startups in 2026, and how do you start it? matter here. The first note should name the trigger, the likely broken workflow, the risk of delay, and one concrete question. Keep it plain.

Compliance is part of credibility. The FTC's CAN-SPAM guidance requires accurate header information, non-deceptive subject lines, clear ad identification where applicable, a valid physical address, and an opt-out mechanism for commercial email senders. That is table stakes. Deliverability also depends on matching message volume to relevance, because a tiny market will burn quickly when every recipient sees the same generic claim.

• One trigger
• One pain hypothesis
• One proof artifact
• One question
• One clean opt-out

Step 5: run founder-led sales until the buyer owns the next step

Sell before you scale. The founder has to hear the buyer hesitate in real time, because the first 10 customers teach the objection map faster than any dashboard will during live sales calls. Use a narrative that starts with the buyer's broken status quo, then borrow the structure in How to build a B2B sales pitch that closes with positioning-led narrative. The demo earns its place after the diagnosis.

Do not chase applause. A useful call ends with a named problem, a named owner, a buying path, an implementation risk, and a next event the buyer would defend to a colleague. I prefer pilots that touch real data or a real workflow, because sandbox praise rarely survives procurement. Done means the buyer accepts work on their side.

• Problem owner confirmed
• Current system named
• Economic pain stated
• Pilot scope written
• Next event owned

What goes wrong

Most founders go broad. They describe a horizontal product to buyers who have different pains, different budgets, different clocks, and different bosses, then blame the channel when the message has no spine. The cure is smaller markets with sharper evidence. A weak no from the right buyer beats a polite yes from a tourist.

Another failure is free labor disguised as traction. If a buyer wants custom analysis, migration help, workflow design, or internal reporting before any commercial commitment, you are funding their indecision. Charge when the work has buyer-specific value. Free can be useful for learning, but it becomes toxic when your roadmap turns into a favors ledger.

• ICP keeps expanding
• Emails sound reusable
• Demos arrive too early
• Pilots lack payment
• Roadmap follows favors

Step 6: convert paid pilots until ten customers share one pattern

Convert patterns, not logos. Your first 10 customers matter because they reveal which pain, promise, sales motion, onboarding work, and buyer role can repeat without heroic founder energy from you. After each close, rewrite the account hypothesis and remove any customer that needed a miracle. The pattern has to be stricter than your gratitude.

Turn the first customer into a sales asset without pretending it proves the market. A short teardown, anonymized workflow map, implementation checklist, buyer-facing calculator, or objection memo can carry proof into the next conversation. For self-serve products, the boundary in When to layer sales onto a self-serve product helps prevent random sales effort from swallowing product signals. Done means the next deal feels less like invention.

• Repeatable pain
• Repeatable trigger
• Repeatable buyer
• Repeatable pilot
• Repeatable handoff

Common questions

How do I get my first 10 customers for my B2B SaaS without an audience?

Start with a narrow ICP and a problem buyers already know is expensive. Build a hand-made account list from public signals, write outreach around a live trigger, run founder-led discovery, and convert paid pilots into repeatable proof. An audience reduces friction later. It should never become the excuse for avoiding specific accounts with visible pain.

Can cold outbound get the first 10 B2B startup customers?

Yes, if the message reads like diagnosis rather than a broadcast. The buyer should see a public trigger, a plausible broken workflow, a reason delay hurts, and one simple question. Cold outbound fails when the founder uses it to test vague positioning at scale. Small volume with sharp evidence beats large volume with generic claims.

Should the first 10 customers be paid?

Usually yes, once the work has buyer-specific value. Free discovery can teach language, but free implementation often hides weak urgency and creates a favors ledger. A paid pilot forces scope, ownership, priority, and timing into the conversation. The price can be modest, but the commitment should be real enough to expose demand.