Hey Automation community! 👋 This took me 2 hours to put together. If you're looking for a proven way to get clients AT SCALE and actually make money for your AI automation agency, then this is for you. I run an AI agency that basically made no money because I had a hard time finding new clients. I tried cold email starting in November and it quickly become one of our most profitable acquisition channels. I knew NOTHING about cold outreach when I started. I learned A LOT along the way (including plenty of expensive mistakes), so here’s everything I wish I had known from day one. If you don't know what cold email marketing is, it's when you send out thousands of emails to potential leads you haven't spoken to before. The goal is for them to book a consult with you where you'll then close on a deal. If you do it badly, it will look like spam and nobody will respond. Do it where you target relevant people ready to buy and offer a lot of VALUE, and you will generate sales. Part 1: Technical Setup Domain Strategy - Buy dedicated domains just for email campaigns — never ever use your main company domain. - Set up DNS records immediately: SPF, DKIM, and DMARC. - Use Google Workspace or Microsoft 365 for better deliverability (roughly $4–6 per account per month). Email Account Setup - Create 1–4 email accounts per domain. - Start slow: 10 emails per account per day, then increase volume by ~10% each day. - Max once warmed up: ~25 emails per account per day. - Example: 4 domains × 3 accounts × 25 emails = 300 emails/day to begin with. IMPORTANT: Always warm up accounts for at least 14 days before ramping up. Extra tips that help a lot: - Add real profile photos and complete the accounts. - Older domains tend to perform better when you can get them. - Set up a custom tracking domain for accurate open/click data. Choosing Your Sending Platform You can do it manually with the technical setup above but it's way easier to buy an email account that's already configured and ready to go. I ran high-volume campaigns using Instantly.ai because it has good deliverability, analytics, and tons of guides on it since it's used by many agencies to get clients. It’s not perfect but probably one of the best for cold email right now. But honestly, your lead list and outreach message matter more.

I thought Day 2 was going to be pretty straightforward. Get Firecrawl connected, learn how MCP servers work, scrape a few pages, check the box, and move on. That lasted about five minutes. Once I had access to real web data inside Claude, I stopped thinking about the tool and started thinking about actual problems I could solve. The setup took less time than expected. What I didn't expect was how fast I stopped thinking about "the MCP server" as the thing I was building, and started just using it. The first real test wasn't a demo page — it was a problem already sitting on my desk. I needed actual product data: door styles, colors, finishes, starting prices, for close to 100 cabinet collections on a supplier's site, for an intake app I'm building for my own home-services business. Before Firecrawl, my plan was to open every page and copy it by hand. Once Firecrawl worked, I pointed it at all 97 pages with a schema telling it exactly which fields to pull, and let it run. One pass, and I had a clean, structured dataset instead of a stack of browser tabs — collection name, brand, door style, color, finish, price range, ready to use instead of retyped by me one page at a time. That's the moment the assignment stopped being "set up a tool" and turned into "the tool just did three hours of my work in one request." The credit system bit me. A search costs 2 credits, and you get 1 back by calling a separate feedback function right after — but only if you remember to. I didn't, more than once, and burned through more of the monthly allowance than I needed to before I noticed the balance dropping faster than it should. A tool only does what you actually tell it, not what you meant. While I was poking at test pages that didn't matter much. Once I pointed it at a real problem with real data going into a real product, every shortcut I skipped showed up as a real cost or a real gap. Has an assignment that was supposed to be "just set this up" ever turned into you solving an actual problem before you even finished reading the instructions?