How to build a prospect list for cold email

March 24, 2026

I sent 200 cold emails from a list I scraped in 20 minutes. 2 replies. Both were “unsubscribe.”

Then I spent 3 hours building a list of 47 prospects manually. Same offer, same template. 9 replies. 4 booked calls.

50 good prospects beat 500 random ones every time. Here’s the process I use now.

Step 1: Define your ICP before you touch a spreadsheet

Most people start building lists by searching LinkedIn for job titles. That’s step 3. Step 1 is knowing exactly who you’re looking for and why.

You need 4 things written down before you open any tool:

1. Problem fit – What specific problem does your prospect have that you solve?
2. Budget authority – Can this person say yes or write the check?
3. Reachability – Can you actually get their email or DM them?
4. Timing signals – What tells you they might need this right now?

For Scouter, my ICP looked like this: marketing leads at creator economy companies with 10–50 employees, who posted a job listing for a partnerships role in the last 90 days. That hiring signal meant they were actively building a creator program – and needed a way to find the right creators.

If your ICP definition fits on a napkin and still feels specific, you’re in the right place. If it’s “anyone who might be interested,” go back and read how to define your ICP for outbound.

Step 2: Find your sources

Where your prospects hang out determines how you find them. Here are the channels I pull from, ranked by signal quality.

Twitter/X

Best for: founders, indie hackers, creators, developer-adjacent roles.

Search for keywords related to the problem you solve. Not your product category – the actual pain. For Scouter, I searched “finding creators is hard” and “creator outreach” instead of “creator discovery platform.”

Follow the conversations, not the bios. Someone complaining about a problem in a tweet is a better prospect than someone with the right job title in their bio.

LinkedIn

Best for: B2B, mid-market, anyone with “Head of” or “VP” in their title.

Use Sales Navigator if you have it. If you don’t, Boolean search still works. The trick is filtering by recent activity – people who posted in the last 30 days are 3x more likely to reply to a cold message than dormant profiles.

Communities and forums

Best for: niche audiences, early adopters, people who self-identify as having the problem.

Slack groups, Discord servers, Reddit threads, indie hacker forums. These are gold because people tell you their problems in their own words. I found 12 Scouter prospects from a single Reddit thread about creator management tools.

Directories and databases

Best for: volume, company-level research, industry mapping.

Product Hunt (recently launched competitors’ upvoters), Crunchbase (funded companies in your space), G2 reviews (people reviewing your competitors). These give you companies. You still need to find the right person at each one.

Tool-assisted discovery

Manual sourcing is high-signal but slow. Tools like Scouter can cut the research phase from hours to minutes when you’re prospecting in the creator economy space. The key is using tools to accelerate the process – not to replace the thinking.

Whatever source you use, track where each prospect came from. You’ll want to know which channels produce the best reply rates later.

Step 3: Build the list

Here’s the spreadsheet structure I use. Nothing fancy. 8 columns.

The “Signal” column is the most important one. It’s the reason this person is on your list instead of the 10,000 other people with the same job title. Recent hire, recent funding, recent complaint on Twitter, recent product launch – something that says “now.”

The “Personal note” column saves you 5 minutes per email later. When you’re doing the research, write down the one thing you’ll reference. A podcast episode they were on. A tweet they posted. A blog they wrote. Capture it while you’re looking at it – don’t make yourself go back.

Build in batches of 25–50. Not 500. You want each batch tight enough that you can write genuinely personalized emails for every single prospect. If you’re copy-pasting the same opener for everyone on the list, your list is too broad.

For the full process on what to look for during research, check how to research prospects before emailing.

Step 4: Verify before you send

Sending to bad emails destroys your sender reputation. A bounce rate above 5% and email providers start routing you to spam. I aim for under 2%.

Use an email verification tool. I run every list through verification before the first send. Takes 10 minutes for 50 emails. Costs almost nothing. Skipping this step costs you your domain’s deliverability.

3 verification rules:

1. Remove catch-all domains – they accept everything, so you’ll never know if the address is real until it bounces
2. Remove role-based addresses – info@, support@, sales@ – these aren’t people
3. Re-verify lists older than 30 days – people change jobs, emails get deactivated

After verification, my 47-person list from earlier became 41 deliverable addresses. 6 bounced during verification instead of bouncing in the inbox. That’s 6 emails that would have hurt my domain reputation for no reason.

Step 5: Segment before you send

Don’t send the same email to everyone on your list. Even a tight list has segments.

I usually split by:

• Awareness level – Do they know the problem exists? Do they know solutions exist? Have they tried competitors?
• Company size – A 5-person startup and a 200-person company need different messaging
• Signal type – Someone who posted about the problem on Twitter gets a different opener than someone you found through a job listing

For my Scouter campaigns, I segment into 3 buckets: active pain (they’ve complained publicly), passive fit (right profile, no signal), and competitor users (they use a competing tool). Each gets a different email template and different subject line.

My numbers

Here’s how list quality affects everything downstream. Same offer. Same sender. Same time period.

The curated list took 3 hours to build. The scraped list took 20 minutes. But I spent more time recovering my sender reputation than I saved on research.

Those 3 hours weren’t just “more work.” They produced a list where every person on it had a reason to be there. That reason showed up in the personalization. The personalization showed up in the reply rate.

The process, compressed

1. Define your ICP – 4 filters, written down
2. Pick 2–3 sources that match your audience
3. Build a list of 25–50 with signal and personal notes
4. Verify every email
5. Segment by awareness level or signal type
6. Send your first batch

Then track what happens. Which sources produced the best reply rates? Which segments responded? Which personal notes got mentioned in replies?

The list is the foundation. A great outreach sequence on a bad list gets silence. A decent sequence on a great list gets conversations.

Build the list first. Build it well.