Back to Tools
GUIDEApril 2026

How to Build a Cold Outreach Pipeline from Scratch in 2026

TC
Trace Cohen
3x founder, 65+ investments, building Value Add VC

Outbound isn't dead — it's just gotten harder to do badly. Here's the exact playbook I give B2B startups to build a cold outreach pipeline that actually books meetings.

My recommended stack: Apollo for data + lemlist for execution

Why Outbound Still Works

Every year someone declares cold email dead. Every year, the companies doing it right keep scaling. The difference in 2026 isn't whether outbound works — it's whether you're willing to invest in the right tools, data quality, and personalization. Spray-and-pray is dead. Targeted, signal-driven outreach is thriving. Here's how to build the machine from zero.

1

Define Your ICP

Before you write a single email, you need to know exactly who you're selling to. Your Ideal Customer Profile isn't “B2B SaaS companies.” It's specific enough that you could filter for it in a database. The more precise your ICP, the higher your reply rates — because your message actually resonates.

🏢

Firmographics

Industry, company size, revenue range, geography, funding stage. These are your baseline filters. Example: “Series A-B SaaS companies, 50-200 employees, US/Canada.”

🔧

Technographics

What tools do they use? If you sell a Salesforce integration, target companies running Salesforce. Tools like Apollo surface tech stack data so you can filter by it.

📡

Buying Signals

Recent funding rounds, job postings for roles your product supports, leadership changes, expansion announcements. These tell you someone is ready to buy now.

Pro tip

Start with your best 5 existing customers. What do they have in common? Reverse-engineer your ICP from reality, not theory.

2

Build Your Prospect List

This is where most pipelines succeed or fail. The quality of your list determines everything downstream. Bad data means bounced emails, wasted sequences, and a burned domain. Good data means you're talking to the right people from day one.

Apollo is my go-to recommendation here. Their database has 275M+ contacts and 73M+ companies with verified emails, direct dials, and intent data. You can filter by every ICP dimension — industry, employee count, revenue, tech stack, job title, seniority, and even intent signals like “researching your category.”

What makes a great prospect list

  • Right personas: Target 2-3 job titles per account. Don't blast the entire C-suite — pick the person who owns the problem you solve.
  • Verified emails: Apollo verifies emails in real-time. Only export contacts with “verified” or “guessed” status — never “unverified.”
  • Intent signals: Layer in buying intent data to prioritize accounts that are actively researching solutions like yours.
  • Manageable size: Start with 200-500 prospects. You can always scale up once your messaging converts.
3

Verify & Enrich Your Data

Even with a great source like Apollo, you should always run a verification pass before loading contacts into your outreach tool. A bounce rate above 3% will start hurting your sender reputation, and once your domain is flagged, it takes weeks to recover.

✉️

Email Verification

Run every email through a verification tool before sending. Apollo has built-in verification, and lemlist also verifies on import. Remove any “risky” or “invalid” addresses. Your target: less than 2% bounce rate.

🧩

Data Enrichment

The more you know about a prospect, the better your personalization. Enrich with company news, LinkedIn activity, recent funding, tech stack, and job changes. This data powers the personalization that gets replies.

Why this matters

Deliverability is the foundation of outbound. You can write the perfect email, but if it lands in spam because your bounce rate killed your sender score, none of it matters. Treat data quality like infrastructure — invest upfront.

4

Set Up Your Email Infrastructure

This is the step most founders skip — and it's the #1 reason their outbound fails. Sending cold emails from your main company domain is like building a house on sand. You need a separate, properly warmed infrastructure.

The setup checklist

  • Buy secondary domains: Get 2-3 domains similar to your main one (e.g., if you're acme.com, buy getacme.com and acmehq.com). Never cold email from your primary domain.
  • Set up SPF, DKIM, and DMARC: These authentication records tell email providers your messages are legitimate. Without them, you're going straight to spam.
  • Warm up with lemwarm: lemlist includes lemwarm, which automatically sends and receives emails from a network of real inboxes to build your sender reputation. Run it for 2-3 weeks before launching campaigns.
  • Create 2-3 mailboxes per domain: Spread your sending volume across multiple inboxes (e.g., trace@, hello@, team@). Keep each inbox under 40-50 emails per day.
5

Write Your Sequences

This is where art meets science. Your sequences need to be personalized enough to feel human, structured enough to scale, and tested enough to improve. lemlist is built for exactly this — multichannel sequences with AI-powered personalization, liquid syntax variables, and native A/B testing.

Personalization at scale

The days of “Hi {first_name}, I saw your company does {industry}” are over. Real personalization references something specific — a recent LinkedIn post, a product launch, a company milestone, a mutual connection. lemlist's AI personalization can generate custom opening lines at scale using real data about each prospect. The result feels like a 1:1 email, but you can send 200 a day.

Go multichannel

Email alone isn't enough. The best sequences combine email, LinkedIn touches (connection requests, profile visits, messages), and sometimes phone calls. lemlist lets you build multichannel sequences in a single workflow. A typical sequence looks like: Day 1 email, Day 3 LinkedIn visit, Day 4 follow-up email, Day 7 LinkedIn connection request, Day 10 final email.

A/B test everything

Test subject lines, opening lines, CTAs, and sequence length. Run every campaign with at least two variants. Small changes in subject lines can swing open rates by 10-20%. lemlist has built-in A/B testing that automatically routes more volume to the winning variant.

Sequence framework that works

Email 1: Personalized pain point + value prop (no ask). Email 2: Case study or social proof. Email 3: Direct CTA for a meeting. Email 4: Breakup email. Keep each email under 120 words. Write like a human, not a marketer.

6

Launch & Optimize

Your first campaign is a hypothesis, not a final answer. The teams that win at outbound are the ones that treat it like a product — ship fast, measure everything, and iterate weekly.

Metrics to track

60%+
Open Rate
Below 40% = deliverability issue
3-8%
Reply Rate
Good for cold outbound
<2%
Bounce Rate
Above 3% = data issue
1-3%
Meeting Rate
% of contacted to booked

When to iterate

  • Low open rates (<40%): Your subject line or deliverability is the problem. Check your warm-up score, test new subject lines, or reduce daily volume.
  • Opens but no replies (<2%): Your messaging isn't resonating. Rewrite your value prop, try a different pain point, or change your CTA.
  • Replies but no meetings: Your CTA is off, or you're reaching the wrong persona. Try asking for a referral to the right person instead of pitching directly.
  • High bounce rates (>3%): Your data quality is poor. Re-verify your list, switch data sources, or reduce the age of your contacts.

The Recommended Tech Stack

After working with dozens of startups on their outbound, I've landed on a two-tool stack that covers everything. Data and execution are different problems — use the best tool for each.

Apollo — Your Data Engine

  • +275M+ contacts, 73M+ companies
  • +Real-time email verification
  • +Intent data and buying signals
  • +Tech stack and firmographic filters
  • +Chrome extension for LinkedIn prospecting
Get started with Apollo

lemlist — Your Execution Engine

  • +Multichannel sequences (email + LinkedIn + calls)
  • +AI-powered personalization at scale
  • +lemwarm for email warm-up
  • +Built-in A/B testing
  • +Email verification on import
Get started with lemlist

5 Common Mistakes That Kill Outbound Pipelines

1

Sending from your primary domain

If your cold emails trigger spam complaints, your entire company's email deliverability suffers — including transactional emails, support, and internal comms. Always use secondary domains for outbound.

2

Skipping the warm-up period

New domains and mailboxes have zero reputation. Sending 100 cold emails on day one is a guaranteed ticket to spam. Warm up for 2-3 weeks minimum, then ramp volume gradually. lemwarm automates this entirely.

3

Writing essays instead of emails

Your cold email should be under 120 words. No one reads a 500-word email from a stranger. Lead with relevance, state your value in one sentence, and end with a clear, low-friction CTA like “Worth a 15-min chat?”

4

No follow-up sequence

Most replies come on email 2, 3, or 4 — not email 1. Sending one email and giving up leaves 60-80% of potential replies on the table. Build 3-4 step sequences with 3-5 day gaps between touches.

5

Targeting too broad

Sending to 10,000 generic contacts will always underperform sending to 500 well-researched ones. Narrow your ICP, personalize your messaging, and scale what works. Volume is not a strategy — precision is.

🚀

Ready to build your outbound pipeline?

Start with the data, nail the infrastructure, personalize your outreach, and iterate fast. These two tools are everything you need.

These are affiliate links that help support Value Add VC at no extra cost to you.