If you’re running B2B marketing, you know drowning in “leads” isn’t the problem—figuring out which ones might actually buy is. If you’ve got a pile of emails or LinkedIn profiles, but no clue who’s worth your time, this guide’s for you. We’ll break down how to segment and score leads with data from Proxycurl, a tool that actually gives you real, up-to-date info on people and companies. I’ll walk you through a process that’s realistic, not pie-in-the-sky “AI will do it all” nonsense.
Let’s get into it.
Step 1: Know What You’re Trying to Find
Don’t start by buying data or setting up scoring formulas. Start by getting clear about your ideal customer profile (ICP). Seriously, skip this step and you’ll waste time chasing the wrong people.
Ask yourself: - What company size buys from you? (Revenue, headcount) - What job titles tend to sign the contract? - Are there industries or locations that matter? - Do you care about funding stage, tech stack, or growth signals?
Write this down. You’ll use it for both segmentation and scoring.
Pro tip: Don’t overcomplicate your ICP. Start with your best customers, not just anyone who could buy.
Step 2: Get Your Leads Ready
You probably have a spreadsheet of emails, LinkedIn URLs, or company names. You’ll want to organize these so you can enrich them with Proxycurl data.
What works: - LinkedIn profile URLs (best for person-level data) - Company domain names or LinkedIn company URLs (for company-level data)
What doesn’t:
- Just names. Too vague—Proxycurl needs something unique to look up.
- Old, spammy lists. If your data’s junk, no enrichment tool will save you.
Clean up your leads list to include: - A column for “Person LinkedIn URL” and/or - A column for “Company LinkedIn URL” or domain
Step 3: Pull in Proxycurl Data
Proxycurl’s main value is turning a dry LinkedIn URL into a living profile: company info, job title, past roles, company headcount, funding, and more.
Here’s how you get the data: 1. Sign up for a Proxycurl account if you haven’t already. 2. Use their API to enrich leads. If you’re technical, you can script this. If not, there are tools or Zapier integrations, or you can even process small batches in their dashboard. 3. For each lead, pull: - For people: Current job title, company, location, work history, skills - For companies: Industry, size, location, funding, tech stack (if available) 4. Add this data to your spreadsheet or CRM.
What works: - Start small. Test a few dozen leads first to see what data you’re actually getting. - Combine person and company data for the best picture.
What doesn’t: - Expecting every field to be perfect. LinkedIn profiles can be out of date, or sparse. - Using free LinkedIn scraping tools—they break, get blocked, or give you junk.
Step 4: Segment Your Leads
Now, use the enriched data to chop up your list into meaningful groups. Segmentation helps you avoid blasting generic emails and lets you focus on the right people.
Common ways to segment: - Company size (e.g., 11–50, 51–200, 201–500) - Industry - Geography - Seniority (e.g., C-level, VP, Manager) - Recent company growth (e.g., just raised funding, lots of new hires)
How to do it: - Use spreadsheet filters or CRM views to sort and tag leads by these fields. - Prioritize segments that match your ICP.
What works: - Start with broad segments. Don’t create 20 categories—start with 3–5. - Use tags or custom fields in your CRM so you can adjust as you learn.
What doesn’t: - Segmenting by “coolness” or vibes. Use hard data. - Over-segmenting. You’re not Amazon, you don’t need micro-audiences.
Step 5: Score Your Leads (Without Getting Cute)
Lead scoring is assigning points to signals that make a lead more or less likely to close. Don’t get lost in the weeds here—simple works.
Basic scoring model:
| Signal | Points | |-----------------------------|--------| | Job title matches buyer | +10 | | Company size in sweet spot | +10 | | Industry matches ICP | +8 | | Located in target region | +5 | | Recent funding event | +7 | | Tech stack matches | +5 | | Junior title (e.g., intern) | -10 | | Company shrinking | -5 |
- Set a threshold: e.g., “20+ points = high priority.”
- Sort your list by score.
What works: - Focus on 3–5 key signals. More than that and you’re guessing. - Calibrate scores over time—see who actually replies or buys.
What doesn’t: - Letting marketing automation “AI” set your scoring for you. It’s rarely smart enough out of the box. - Relying on activity signals (who opened an email) instead of hard data.
Step 6: Actually Use the Segments and Scores
This is where most teams drop the ball—they segment and score, then…nothing happens. Here’s how to make it count:
For high-score leads: - Prioritize for sales outreach—call, email, connect on LinkedIn - Personalize your messages with what you know (e.g., “Congrats on your recent funding”)
For mid-score leads: - Add to targeted nurture campaigns - Watch for new signals (maybe they change jobs or companies grow)
For low-score leads: - Park them for later. Don’t blast them with everything—keep your brand in good standing.
What works: - Regularly review who’s moving up or down in score. - Share feedback between marketing and sales—who’s actually converting?
What doesn’t: - Ignoring the data. If your high-score leads never reply, revisit your scoring. - Setting and forgetting. Segments and scores should evolve.
Step 7: Keep It Simple and Iterate
You don’t need a giant martech stack or fancy dashboards to get value from Proxycurl data. The basics—clean data in, clear ICP, simple scoring—are what actually move the needle.
A few honest tips: - Don’t expect enrichment to magically fix bad lists. - Your first scoring model will be wrong. That’s fine. Tweak it. - Ask sales if the “best” leads are actually buying—if not, adjust your criteria. - Focus on actionable data, not vanity metrics.
Shortcuts that don’t work: - Buying massive lists and enriching them all. Quality > quantity. - Trusting any tool that promises “AI-powered intent” without explaining where the data comes from.
Bottom line:
Segmenting and scoring leads with Proxycurl data isn’t rocket science, but it does take a little discipline and a lot of iteration. Keep your process simple, use real data, and don’t overthink it. The goal: spend more time on leads who might actually buy, and less on those who never will. That’s it. Good luck!