Mailshake Prospecting Checklist: 6 Steps to an Agent-Native Workflow (With Reverse Email Lookup)

I coordinate outbound at a B2B SaaS company. In four years, I've led roughly 40 campaign launches—and more than a dozen of them were emergency turnarounds. The pattern, more often than not, is the same: a sales leader appears at 2 PM, the quarter ends in 10 days, and they need 1,000 targeted conversations started by Friday.

When that happens, the instinct is to grab an email tool, upload the handiest list, and start firing. That's backwards—and it's why rushed campaigns tend to produce volume without replies.

This checklist is for teams that need to launch a Mailshake prospecting workflow in days, not weeks. It covers the order that actually works: API company data, the account list, reverse email lookup, the campaign, compliance, and the agent layer that runs on top. Six steps. Step one doesn't involve Mailshake at all.

Before you begin, open the Mailshake official homepage (mailshake.com) and read the integration list—especially Google Sheets and HubSpot. The official site also lets you book a Mailshake demo. The demo is worth seeing, but it won't tell you how to structure your data. That's what this checklist is for.

The six-step checklist

Step 1: Define the account universe using API company data

The natural instinct is to start by collecting emails. It's a blind spot: teams obsess over subject lines and open rates, then upload a list of domains that were never screened. The list is the campaign. If the companies don't fit your ICP, no template saves you.

API company data—industry, employee count, funding stage, tech stack—is the filter. Before you write a single line of copy in Mailshake, pull the account universe from an enrichment API and apply your ICP filters. Here's a simple checkpoint: you should be able to describe the target account in one sentence. “Series B+ fintech companies, 50–500 employees, using Salesforce, in North America.” If you can't say it in one sentence, you're not ready to send.

The data doesn't need to be exotic. A basic enrichment API with the right filters beats a premium list with none—what matters is that the attributes you filter on match the ones your best customers actually share.

Step 2: Build the account list in Sheets or HubSpot first

This step feels like a detour, but it's the one that makes everything else possible. Pull your enriched accounts into a Google Sheet or HubSpot, with each firmographic attribute in its own column. Then connect that source to Mailshake.

Why an intermediary instead of loading contacts straight into the campaign? Because the Sheet is the single source of truth. In an agent-native workflow, your AI agent, your SDRs, and your Mailshake sequences all read from the same file. When an account changes—new funding round, new headcount—you fix it once, and every downstream system sees the update. Agent-native doesn't mean a chatbot bolted onto the tool. It means the workflow is structured so an agent can run it: clean data in, clean actions out.

Step 3: Run reverse email lookup to attach the right contact

Now you have 500 screened accounts. You need a person at each. That's where reverse email lookup fits: feed the company domain into a lookup tool, get the decision-maker's email, and log it back into the Sheet or CRM.

The order is where rushed launches tend to fall apart. The reverse lookup doesn't qualify the account—the company data in Step 1 does. If you run lookups before filtering, you'll spend credits on a thousand people at companies that were never going to buy. Accounts first, contacts second. The email list is an output of good targeting, not the starting point.

Mailshake's lead generation features—email verification, the Sheets/HubSpot sync, and the sequence tools—cover the pieces that eat the most time in a rushed launch. Verification alone prevents a long list of deliverability problems later.

Step 4: Build the campaign on firmographic personalization

Now the part people assume is step one: the conversation itself. In Mailshake, create a new campaign and add your sequence. Email 1, a follow-up a few days later, another email, maybe a LinkedIn touch if your stack includes Sales Navigator. Keep it short. A 10-email sequence is a confession that the first nine didn't work.

Personalization should come from the company data, not just the first name. “Noticed you're at a Series B fintech running on HubSpot—” works because it's specific and verifiable. That sentence only exists because Step 2 happened.

This is where the Mailshake demo becomes genuinely useful. A typical demo shows you templates, cadence builders, and reporting—the polished stuff. Ask instead about the plumbing: how do custom variables map to the columns in your Sheet? What happens to replies when a prospect answers? Where does the handoff occur? The demo looks good either way. The honest ones show you the wiring.

Step 5: Verification and compliance guardrails

Run verification before the send. Obvious on paper, easy to skip in an emergency. We skipped it once in 2024 during a rushed launch and spent the next month repairing domain reputation.

Then the compliance layer. Per FTC guidance (ftc.gov), commercial email needs accurate header information, a real physical address, and a working opt-out mechanism. This isn't a gray area. An unsubscribe link that doesn't work is a violation, not an oversight.

To be fair, compliance setup adds a small overhead. In practice it takes about 20 minutes in Mailshake. The alternative—a blocked domain or complaint-driven shutdown—ends the campaign anyway. There's no time saved by skipping it.

Step 6: Put the agent on top of the data pipeline

Now the real answer to the API company data question: in an agent-native prospecting workflow, company data is the memory of the agent.

An AI agent can't know whether an account is a fit unless the data layer tells it. In a working setup, the agent:

• Queries the company data API on a schedule to refresh the account list
• Reads replies as they land in Mailshake
• Scores interest from the reply text and the firmographic context
• Routes hot conversations to an account executive and moves the rest to nurture

Mailshake is the delivery engine. The agent runs on top of the data pipeline. If the pipeline is clean, the agent's decisions are clean. If it's garbage—unscreened domains, stale titles, no firmographic filters—the agent will confidently hand your AEs a list of people who were never going to buy. That risk is worse than any bounce rate.

When the order is right, it's fast. In October 2025, we had 36 hours to turn a 1,200-account list into a live campaign. We spent the first 6 hours on data—API pulls, enrichment, Sheet columns, verification—and the last 4 writing copy and building the sequence. The sending itself took minutes. Every rushed launch that failed before that date had spent its time in the wrong order. A lesson learned the hard way.

Not every team needs the agent layer. If you're running 30 outbound touches a month, a spreadsheet and a human are fine. But once you're scaling past what a human can triage, the agent layer stops being optional.

What to watch out for

The “deliverability first” distraction

The question I hear most often is, “What's your deliverability?” The question I wish more teams asked is, “Where does our company data come from, and how fresh is it?” Deliverability gets emails to the inbox. The data layer decides whether the inbox belongs to someone who can buy.

The static list trap

If your Sheet isn't connected to anything, the enrichment is a one-time photograph. Company data changes—headcount, tech stack, funding. An agent-native workflow refreshes on a schedule. A manual workflow forgets, and next quarter the list is stale. Once it's stale, the personalization stops being true, and the replies stop coming.

Expecting the tool to replace judgment

Mailshake doesn't replace SDRs. It replaces the repetitive work—sending, following up, logging, handoff. The judgment still lives in the ICP and the offer. If the one-sentence description from Step 1 is wrong, the tool just makes the wrong thing louder.

The hidden-cost surprise

Granted, this workflow has more pieces than buying a list and hitting send. There's a data API subscription, lookup credits, and a Mailshake plan. All of it shows up on the invoice—and I prefer it that way. A vendor that hides fees until later, or keeps the limits of its data layer vague, costs more in the end. Not always in dollars; often in wasted hours and burned domain reputation.

The cheap list is the expensive one. The tool that's transparent about its limits, and lets you inspect the wiring, is the one that actually helps when the deadline is real.