How to Start Outbound Sales for a Small Business

Starting outbound sales from zero needs less setup than most advice implies: a clear picture of who you’re targeting, a way to find and verify their contact details, and a short, honest first email. Sequencing software, elaborate tooling, and a big list all come later, if they turn out to be needed at all.

1. Define who you’re actually targeting

Write down, specifically: company size, sector, geography, and the situation that makes a company a fit right now (not just in theory). “B2B companies” isn’t a target market, it’s the absence of one. If you can’t describe your ideal prospect in two sentences, that’s the actual first task, before any list-building starts.

2. Build a small, real list before a big one

Ten to twenty companies that genuinely fit is enough to start. See how to find prospects without a bulk database for where to source it without paying for a large data subscription upfront.

3. Set up the minimum technical foundation

Before sending anything, confirm the domain you’re sending from has SPF, DKIM, and DMARC configured, and plan to ramp volume gradually rather than sending a large batch on day one. See cold email deliverability basics for the full checklist. Skipping this step is the most common reason a first campaign underperforms for reasons that have nothing to do with the writing.

4. Research before writing

For each prospect on the list, confirm the right contact and find one true, current reason to reach out. See how to research B2B prospects. This step takes longer than the writing, and skipping it is the difference between outreach that gets replies and outreach that gets ignored.

5. Write short, and send a small batch first

Use a simple, fixed structure (see how to write a research-based cold email), and send to a small batch first, ten to twenty emails, before scaling up. This lets you catch anything wrong (a broken link, an unclear ask, a deliverability issue) before it affects a larger list.

6. Track replies, not just sends

The number that actually matters is reply rate, and within that, the fraction of replies that turn into a real conversation. Sending volume is easy to increase and easy to mistake for progress. It isn’t progress if nobody’s replying.

What to add later, once the basics are working

• A larger, ongoing list, once you’ve validated your targeting on a small one.
• Sequencing or follow-up automation, once you know the first email works.
• More hands, whether a hire, a configured self-serve platform, or a managed service, once volume outgrows what one person can research and write personally. See should a startup hire an SDR or outsource outbound.

The five steps, in order

1. Write an ICP you can verify from outside. Industry, size band, geography, business model, stage. The test: can you look at a website and decide in under a minute whether it qualifies? If not, keep narrowing.
2. Build a list of fifty accounts. Fifty, not five thousand. Large enough to show a pattern in the replies, small enough to research each one properly.
3. Set up sending properly before the first email. A separate domain, SPF, DKIM, DMARC, and a warmed mailbox. This takes an afternoon and prevents a category of problem you cannot undo.
4. Research and write to all fifty individually. One verified address and one current signal per account. Accounts with no signal get dropped rather than padded.
5. Read every reply before scaling anything. Including the declines, especially the declines.

Why fifty

Fifty well-researched accounts produce enough replies, including negative ones, to see a pattern, and few enough that you can hold all of them in your head at once. That second property is what makes the exercise informative rather than merely productive.

Five thousand poorly targeted accounts produce a number, and the number tells you nothing you can act on. It also burns a domain and a market segment you may want back later, which makes it strictly more expensive than doing less.

What you are actually trying to learn

• Which segment replies. Your first ICP will be wrong. The first fifty tell you which direction it was wrong in, which is the most valuable output of the whole exercise.
• Which objection recurs. A repeated objection is positioning feedback arriving through a sales channel. It usually means the offer needs changing, not the email.
• Which framing earns a reply. Record the message that worked verbatim. Paraphrasing it later loses the part that worked.

What not to do first

• Do not automate before a message has earned replies manually.Automation multiplies what you already have, including nothing.
• Do not buy a list. Unknown provenance, unknown age, no timing information, and a compliance problem in the UK and EU.
• Do not send from your main domain. That is the mail your business depends on.
• Do not judge it on meetings booked in month one. Meaningful, repeatable pipeline from a cold start is months rather than weeks. Judge the first cycle on what you learned.

If you are the founder doing this yourself, which at this stage you should be, the founder-led outbound playbook covers how to run it as a defined block and when to hand it off.