How can I deeply research prospects and personalize cold emails when I find it hard to get information? How deep should research go, and can you give a quick guide?
Don't personalize the whole email—personalize only small snippets using enhanced variables. For example, use a template like 'Hey {first name}, love {thing we have in common}, fun to see you mention that on {platform} where I got it.' Then continue that vein to your offer. If AI writes the whole email, it will likely produce something stupid and obvious as AI‑generated. Keep the AI constrained everywhere and feed it only small variables (enhanced variables) with examples, letting it spin the initial info into a tightly woven, human‑sounding message. For research, use Claude to look up the person, compile company info, and grab all social media profiles; if you can't, use scrapers or similar tools.
It’s a great time to build personalization flows because new entrants in cold email are higher than ever. Cold email had a big renaissance, AI personalization hit, and it still outperforms other marketing methods when run competently. Personalization boosts ROI, but you need to go beyond simple icebreaker lines—constrain the AI output, use your copywriting to write templated bits like “I love that you’re doing X. It’s refreshing to see that you’re doing that. Figured I’d reach out.” If you don’t constrain, AI‑generated icebreakers feel flamboyant and are seen as AI‑written, hurting perceived personalization. Instead, personalize the specific high‑ROI parts of the email (not the whole thing or just one line). Examples: paraphrase three interesting things from a prospect’s podcast, or pull key quotes from their content. You can run thousands of these personalized emails a day at a token cost (~$20/month). There’s massive untapped potential—cold email volume will soon far exceed what most people are doing.
First, a 2 % reply rate isn’t terrible, but there’s room for improvement. Personalizing only the first line of an email is no longer enough—everyone does that now. Think of personalization as a foundational strategy rather than a tactical tweak. The core idea is to build an icebreaker that shows you’ve actually researched the prospect, not just scraped their name.
Start with an icebreaker that combines a genuine compliment with a paraphrased reference to something they posted on LinkedIn, Facebook, or their website. For example, “Hey Nick, I deeply respect your work, especially your recent post about AI‑driven marketing.” To create the paraphrased part, pull a specific detail from their content, feed it to AI, and ask it to rewrite it in your own voice. This makes the prospect feel you took the extra step to understand them.
Next, add an identification line that tells who you are and why you’re reaching out. Keep it short: “I’m a marketer who helps agencies like yours generate $1‑5 k of monthly revenue using AI‑enhanced outreach.” Mention a concrete credential or past success—e.g., “When I ran PR campaigns for Forbes‑level publications, I saw 20 % reply rates.”
Follow the icebreaker and ID with social proof: cite a recent win that’s relevant to the prospect’s industry. For instance, “Last week I delivered $8,390 in revenue in two months for a Sydney‑based marketing agency similar to yours.”
Finally, present a clear offer: “I’m confident I can replicate—or even improve—those results for you.”
Test multiple variations. Write at least ten different icebreakers and identification lines, then pit them against each other to see which combination yields the highest response. By focusing on deep research, personalized paraphrasing, concise self‑identification, and strong social proof, you’ll move well beyond the 2 % baseline and achieve much higher reply rates.
Should you hyper-personalize every one of your 500 leads in cold email? Yes, you should. And you should use AI to help you do so. Hyper-personalize the first five or 10 yourself, and then just have AI pluck out the different patterns involved in your personalizations and try to replicate them. Then you'd be pretty good. Always try doing it yourself at least once — if you do it yourself at least once, then you'll have a good template and baseline for AI to work off of.
Here’s how I’d build a fully autonomous reply bot for a cold‑email campaign that sends 10 000+ messages a day. I’d start with a cheap inference platform such as N8N and create an AI agent inside it. The agent would have access to a knowledge base about the company and a library of proven reply templates collected from the past few months.
When a new inbound email arrives, a workflow would trigger that (1) pulls the entire conversation history with that prospect, (2) checks the knowledge base to confirm the company context, and (3) scans the template library to find the closest match for the desired outcome (e.g., booking a meeting). The selected template would be populated with any dynamic details, and the AI would generate the final reply while staying as close to the template as possible to avoid sounding overly AI‑generated.
The idea is to give the AI minimal freedom—just enough to fill in variables—so you retain control over tone and quality. I originally built this in N8N, but once you scale to 100 000 emails per day you can replace the visual builder with lightweight Python scripts running on a serverless platform like Modal, listening to webhook events from your email service.