Outbound email is still one of the highest-ROI channels for B2B lead generation — if you do it right. But doing it right means personalization at scale, smart follow-up sequences, and clean list management. That's a lot of manual work unless you automate it. Over the past year I've built Claude Code workflows that handle most of the heavy lifting in my outbound email systems, and they've consistently outperformed the template-based tools I used before.
This post walks through how I'm using Claude Code to automate outbound email in 2026 — the workflows that work, the mistakes to avoid, and the compliance considerations you can't ignore if you're operating in Canada.
Why Claude Code for Outbound Email Automation
There are a lot of tools that claim to automate outbound email. Lemlist, Instantly, Woodpecker — they all do a version of the same thing. You upload a list, pick a template, and they send emails on a schedule. For simple campaigns, they work fine. But they all hit the same ceiling: the personalization is shallow, the logic is rigid, and integrating them with your CRM or enrichment tools requires a separate no-code platform like Zapier.
Claude Code lets you build the entire stack yourself. You own the data pipeline, the email copy generation, the sending logic, and the follow-up rules. That means you can:
- Pull live data from LinkedIn, company websites, or news APIs to personalize every email
- Write dynamic email copy that adjusts tone and structure based on recipient role or industry
- Build conditional follow-up sequences that change based on opens, clicks, or replies
- Integrate directly with Gmail API, SendGrid, or your CRM without middleware
The tradeoff is you need to build it. But once you do, you control everything — and the cost per send is close to zero.
The Core Workflow: List Enrichment to First Send
Here's the basic pipeline I use for most outbound campaigns. It starts with a raw contact list — usually scraped from LinkedIn Sales Navigator or pulled from a trade association directory — and ends with personalized emails in the recipient's inbox.
Step 1: Enrich the Contact Data
A raw list usually has a name, company, and maybe a job title. That's not enough to write a good cold email. I use Claude Code to enrich each contact with:
- Company size, industry, and recent news (scraped from the company website or a news API)
- LinkedIn profile summary (if publicly available)
- Technologies they use (from BuiltWith or similar tools)
- Any mutual connections or shared groups
This enrichment step takes about 2–3 seconds per contact and runs in the background. For a list of 500 contacts, it's done in under 20 minutes. The output is a JSON object for each contact with all the context I need to write a relevant email.
Step 2: Generate Personalized Email Copy
Once the data is enriched, I feed it into a Claude Code prompt that writes the email. The prompt includes:
- The recipient's name, title, company, and industry
- A specific pain point relevant to their role (pulled from a predefined list based on job title)
- A reference to something recent — a company milestone, a LinkedIn post, or an industry trend
- A clear, low-friction CTA (usually a question, not a meeting request)
The key to making this work is the prompt structure. I don't ask Claude Code to "write a cold email." I give it a formula: opening hook (1 sentence, specific to them), pain point (1–2 sentences, tied to their role), how I solve it (1 sentence, no jargon), and a question CTA. The output is always under 100 words and reads like I wrote it manually.
For a Vancouver-based marketing agency I built this for, personalized email generation cut their outbound prep time from 3 hours per 100 emails to about 15 minutes — and reply rates went up 40% because the emails felt more relevant.
Step 3: Send via API
Once the emails are generated, they get queued in a CSV or JSON file. A separate script connects to the Gmail API (or SendGrid if you're sending higher volume) and sends them in batches. I usually stagger sends over a few hours to avoid triggering spam filters.
The sending script also logs every send: recipient email, timestamp, subject line, and a unique tracking ID. This makes it easy to tie replies or opens back to the original contact record.
Automating Follow-Up Sequences with Conditional Logic
Most cold emails don't get a reply on the first send. A good follow-up sequence can double your reply rate, but only if it's contextual. Sending the same "just bumping this" message to everyone is lazy and it shows.
With Claude Code, I build follow-up sequences that branch based on behavior:
- No open after 3 days: Send a different subject line and a shorter email (2–3 sentences max)
- Opened but no reply after 5 days: Send a value-add follow-up — a relevant case study, article, or tool recommendation (no ask, just helpful)
- Replied but didn't book a call: Move them to a nurture sequence with monthly check-ins
The logic for this lives in a simple script that checks the status of each contact every 24 hours and triggers the appropriate follow-up. It's not complicated code, but it would be impossible to replicate in a traditional email tool without paying for their enterprise tier.
Compliance: CASL and Outbound Email in Canada
If you're sending outbound email from or to Canada, you need to follow Canada's Anti-Spam Legislation (CASL). The law is stricter than the U.S. CAN-SPAM Act, and the fines are real — up to $10 million for businesses.
The key rules:
- You need consent — either express (they opted in) or implied (you have an existing business relationship or they made their contact info publicly available in a relevant context)
- Every email must include an unsubscribe link and your physical mailing address
- You must honor unsubscribe requests within 10 business days
For most B2B outbound, implied consent is the relevant category. If someone's LinkedIn profile says "Head of Marketing at XYZ Corp" and their email is publicly listed, you can email them about marketing services. But you still need the unsubscribe link and you still need to stop if they ask.
I built CASL compliance directly into my Claude Code email workflows. Every generated email includes an unsubscribe footer with a link to a simple form that adds the email to a suppression list. The suppression list gets checked before every send. It's not optional — it's part of the pipeline.
What This Workflow Can't Replace
I want to be clear about the limits here. Claude Code is great at automating the mechanics of outbound email, but it's not a replacement for strategy. You still need to:
- Define your ideal customer profile and build a quality target list
- Write a compelling value proposition that resonates with your audience
- Test subject lines, CTAs, and messaging angles to see what works
- Handle replies and book meetings — that's still a human job
Automation amplifies what's already working. If your messaging is weak or your list is bad, automation just means you'll annoy people faster. Start with a manual campaign, get your reply rate above 5%, and then automate.
Getting Started: The 48-Hour Build
If you want to build a version of this for your own outbound campaigns, here's the fastest path:
- Day 1 morning: Build the list enrichment script — pull company data and LinkedIn profiles for your target contacts
- Day 1 afternoon: Write and test the email generation prompt — aim for 10 test emails that you'd be willing to send yourself
- Day 2 morning: Set up the sending script — connect to Gmail API or SendGrid and send your first batch of 50 emails
- Day 2 afternoon: Build the follow-up logic and the unsubscribe handler
That's a working system in two days. From there, you can add complexity — A/B testing, reply detection, CRM integration — but the core pipeline is enough to start generating replies.
If you're running a Vancouver-based agency or consultancy and want to see how this could work for your specific use case, I walk through live examples on this getting started guide. And if you have questions about CASL compliance or integration with your existing CRM, the FAQ page covers most of the common scenarios.
Four Takeaways
- Enrichment before personalization: The quality of your email copy depends on the quality of your data — invest time in the enrichment step
- Test manually first: Don't automate a campaign until you've validated the messaging with at least 50 manual sends
- CASL compliance is non-negotiable: Build unsubscribe handling into your workflow from day one — it's not worth the risk
- Follow-ups matter more than the first send: Most replies come after the second or third touch — build conditional follow-up logic, not generic bumps
The tools are here. The question is just whether you want to keep paying per-send fees to a third-party platform or build a system you control. If you're sending more than 500 emails a month, the economics make the choice obvious.