LinkedIn Scraping, Email Verification & Sales Cadence: A Phantombuster B2B Checklist

I've been running B2B prospecting operations for five years. In that time, I've personally made (and documented) seven significant mistakes, totaling roughly $6,300 in wasted budget. That number doesn't include the slower cost: a damaged sender reputation, annoyed sales reps, and a team that stopped trusting the lead list. This checklist is what I use now. It's not a theory piece. It's the operational order I follow when I set up LinkedIn scraping, email verification, and a sales cadence for a new campaign.

Who This Checklist Is For

Use this if you're a B2B sales team or revenue operations person who:

• Has a LinkedIn Sales Navigator account and a tool like Phantombuster but is not sure how to turn exports into replies.
• Has bought a cheap data tool or plan because the price looked great, then watched the campaign fail for reasons that had nothing to do with the tool.
• Needs a simple, repeatable process for scraping, verification, and outreach that doesn't depend on one person remembering everything.

This is a seven-step checklist. The first four steps take about an hour. The last three are ongoing discipline. If you skip the boring parts, you're gonna spend the same money I wasted.

The Checklist: 7 Steps I Use Before Every Prospecting Campaign

Step 1: Define the person, not the list size

Before you open a scraper, write the exact search query you'd type into LinkedIn Sales Navigator. 'Anyone with VP in their title' is not a query. 'VP Sales at B2B SaaS companies with 50-500 employees in the UK' is closer. Add exclusions: existing CRM contacts, current customers, competitors, and anyone already in a sales cadence.

It's tempting to think that LinkedIn scraping is just 'run a scraper, download a CSV, and import it into a cadence.' But the CSV is not the deliverable. The clean, verified, segmented list is the deliverable.

Step 2: Verify email addresses before you enrich them

This is the step I used to skip. In September 2022, I exported about 4,000 contacts from Sales Navigator, skipped verification because I was in a hurry, and launched a 4-touch email cadence. The bounce rate was bad enough that our sending domain got flagged. We didn't notice until a legitimate proposal email landed in spam. That one issue cost us a week of fixes and a deal I'd rather not remember.

So, what is email verification and when should a B2B sales team use it? Verification checks whether an email address can receive mail before you include it in a campaign. It usually checks syntax, domain mail exchanger (MX) records, and often runs an SMTP check with the receiving server. It does not tell you if the person is a good fit. It does not tell you if they are ready to buy. It reduces the chance you'll send to an address that doesn't exist or won't accept messages.

Use email verification whenever:

• The list came from a scrape, event attendee list, or a third-party source.
• The list is more than 30 days old.
• You are about to send to more than a few hundred contacts.
• A new campaign is going to use a brand new sending domain.

Also, keep your email basics legal. Per FTC guidance (ftc.gov), commercial email must include a clear opt-out and a valid physical address. That's not a deliverability trick, it's the legal floor.

Step 3: Scrape only the fields you can actually use

When I audited our wasted data, I found we were exporting more than 40 columns and using about eight. The other columns were not free. They consumed time, confused the CRM, and created fake confidence. Scrape what the cadence needs: first name, last name, title, company, LinkedIn URL, email, and one signal for personalization. If a campaign uses a different signal, add it then.

Step 4: Segment by signal strength

When I compared our Q1 and Q2 email results side by side—same sending tool, same cadence, different data quality—I finally understood why segmentation matters more than volume. The top-tier segment produced replies at more than double the rate of the raw list. The raw list still had some value, but it needed a different sequence.

I sort every verified list into three buckets: strong fit, mid fit, and broad. Strong fit gets a personalized email plus a LinkedIn connection attempt. Mid fit gets a shorter sequence. Broad fit only gets a monthly check-in, if we send anything at all. This approach saved us a ton of time and stopped the one-size-fits-all cadence.

Step 5: Use the right Phantombuster tool for the action

Phantombuster is a no-code automation platform. It has a tool for scraping LinkedIn Sales Navigator results and a tool called LinkedIn Network Booster for sending connection requests. I use both, but I use them for different actions. The scraper builds a list. The Network Booster starts a connection. They are not interchangeable.

Here's a setup that works well for me:

1. Export a targeted list with the LinkedIn scraper.
2. Verify the emails before you do anything else.
3. Upload profile URLs to LinkedIn Network Booster.
4. Set conservative daily limits and a short, human message.
5. Export accepted connections and add them to the email sequence.

I won't promise LinkedIn automation is risk-free. No one can promise that. But you can reduce risk by keeping request volume moderate, personalizing the first line, and leaving some effort for manual follow-up. You're responsible for respecting platform terms.

Step 6: Build a sales cadence with a reply path, not a send schedule

A sales cadence is a planned sequence of touches: connection request, email, follow-up, maybe a phone call. The mistake I see in our own history is building cadences around what we wanted to send, not around what the buyer would tolerate.

Rule I use now: no verified list, no cadence. No suppression list, no send.

A simple cadence looks like this: Day 1 LinkedIn connection request. Day 2 first email. Day 5 first follow-up. Day 9 second follow-up. Day 14 a breakup email that says 'I'll close your file unless you tell me otherwise.' Every email should have a task assigned to the sales rep if the person replies. If no one is assigned to reply, don't send it.

Step 7: Measure cost per reply, not cost per lead

This is where the value-over-price argument shows up. In my experience managing prospecting budgets, the lowest quote has cost us more in most cases. A $30/month difference between tools is nothing compared to a team wasting a week on dead contacts.

The number to watch is cost per qualified reply. Let's say a campaign costs $300 in tools and verification. If it produces 30 qualified replies, the cost per reply is $10. If a free data source produces 5 replies, it's not free. It costs you sender reputation and rep time. That's the hidden cost.

Phantombuster Pricing 2024: The Hidden Cost Mistake

When I reviewed Phantombuster pricing in 2024, I almost made the classic buyer mistake: I compared the lowest monthly plan on a spreadsheet and ignored what each operation actually costs. Phantombuster pricing is based on operations, not on the number of users. That's a meaningful difference.

LinkedIn Network Booster runs consume operations. A multi-step sequence can consume a lot of operations. If you buy a plan with a small operation limit, you'll hit the cap before the campaign finishes. Then your choices are wait, upgrade, or lose momentum. All of those have a cost.

Don't hold me to the exact operation counts—check Phantombuster's current pricing page. My advice is to pick a plan with roughly twice the operations you think you need. You can adjust next month. But starting too small saved me $20 one year and cost me about $700 in lost time and deliverability fixes. That $700 came from the exact mistake I'm avoiding now: choosing the lowest price instead of the right capacity.

To be fair, a smaller plan is fine for a test. But a test should be small in scope, not small in capacity. Run 3-5 small automations, review the results, then scale.

Common Mistakes I Made So You Don't Have To

Some of these are embarrassing. I'm sharing them because they're cheaper to learn from somebody else's spreadsheet.

Mistake 1: Using a scraped list older than 30 days

I once sent a 1,200-contact campaign from a list that had been sitting in Google Sheets for six weeks. The bounce rate was high. The sender reputation took days to recover. I now re-verify any list that is older than 30 days before a major send.

Mistake 2: Not creating a suppression list

If someone already replied, unsubscribed, or is a current customer, they should never appear in a new scrape. It sounds obvious, but I've done it. It erodes trust and makes your metrics look better than they are. I now keep a suppression list that is imported into every scrape and every send.

Mistake 3: Running LinkedIn Network Booster without testing the message

I once launched a connection request sequence with a question that was too long. I didn't test it on my own profile first. The message looked cluttered on mobile, and the acceptance rate dropped. Test every automation on a small sample before you scale.

Bottom Line

The 'you need 5,000 new leads a month' thinking comes from an era when cold email was less protected and mailbox filters were less aggressive. That's changed. A smaller list of verified, relevant contacts will almost always beat a big list of raw names.

When I started in sales ops, I thought the job was to make sending cheaper and faster. It isn't. The job is to make the next conversation more likely to happen. That means scraping with intention, verifying addresses before they enter a cadence, and using tools like Phantombuster with a clear operational plan.

I've made the expensive mistakes. We've caught 47 potential errors using this checklist in the past 14 months. That's 47 mistakes that didn't reach a prospect's inbox. Follow the same process, and I'm guessing you'll catch yours before they get expensive too.