The techniques we use to make automated outreach feel genuinely human, and the data showing why it works.
Everyone's inbox is full of terrible cold emails. You know the ones. "Hi [First Name], I noticed your company is in the [Industry] space and thought you might be interested in our solution..." Delete. Straight to the bin.
The reason most cold emails fail isn't because cold outreach doesn't work. It's because the emails are lazy. They're templates with a name swapped in, and the person receiving them can spot it immediately. Nobody wants to feel like they're on a mailing list.
But here's the thing. Genuinely personalised outreach works remarkably well. When someone receives an email that references something specific about their business, mentions a challenge they're actually facing, or ties into something they recently posted about, the response rate goes through the roof. The problem has always been that real personalisation takes time. A lot of time.
That's where AI changes the game.
When people hear "AI written emails" they picture ChatGPT vomiting out generic paragraphs full of phrases like "in today's rapidly evolving landscape" and "leverage synergies." And honestly, if that's what you're doing, you're better off not automating at all.
The trick isn't asking AI to write an email from scratch. It's giving AI the right data about a specific prospect and asking it to weave that data into a proven messaging framework. The AI isn't being creative. It's being precise.
Good personalisation starts well before the email gets written. The system needs data to work with, and the quality of that data determines the quality of the output. Here's what our systems typically pull in for each prospect:
All of this gets collected automatically. No human has to Google anyone or scroll through LinkedIn profiles. The system does it in seconds.
Once the research is done, the AI writes the email using a framework we've developed and refined over hundreds of campaigns. The framework has a few key principles:
The first line of the email should be about the prospect, not about you. "I saw your team just expanded into the US market" hits very differently to "We're an automation agency based in Yorkshire." Nobody cares about you until they know you understand them.
The personalisation should naturally lead into why you're reaching out. If you mention that they recently hired three new SDRs, the logical next step is talking about how to get those new hires productive faster. The transition from "I noticed this about you" to "here's how we can help" should feel effortless, not forced.
Nobody reads long cold emails. Our emails are typically four to six sentences. That's it. The goal isn't to explain everything you do. It's to earn a reply. All the detail comes later in the conversation.
This is the part most AI setups get completely wrong. The default output from most language models sounds like a corporate brochure. We've spent a lot of time training our systems to write the way real salespeople talk. Short sentences. Contractions. The occasional rhetorical question. No jargon unless the prospect uses it themselves.
Subject: Quick question about your outbound
Hi Sarah,
Noticed you've just brought on two new BDRs. Congrats on the growth. Quick question: are they doing prospect research manually or have you got something in place to handle that?
Asking because we built a system for [similar company] that cut their research time from 20 mins per prospect to about 30 seconds. Their team went from 8 meetings a week to 22 within the first month.
Worth a quick chat to see if something similar would work for your lot?
Cheers,
[Name]
That email took about four seconds to generate. It references a real event (the new hires), ties it to a relevant problem (manual research), and offers proof that the solution works (specific results from a similar company). It reads like something a thoughtful salesperson would write, except no salesperson would have the time to write something this targeted for every single prospect on their list.
We've been tracking the performance of AI personalised emails against traditional template based outreach across our client base. The differences are significant:
The maths is straightforward. If you're sending 500 outbound emails a week with a 3% reply rate, that's 15 replies. With AI personalisation pushing that to 10%, you're looking at 50 replies from the same volume. That's not an incremental improvement. That's a fundamental shift in what your outreach programme can deliver.
And the best part? It's completely scalable. The system that personalises 50 emails works exactly the same way for 500 or 5,000. The quality doesn't drop as volume increases because every email gets the same depth of research and the same quality of writing.
The era of spray and pray outreach is over. The businesses that win are the ones that make every prospect feel like the only person they're talking to, even when they're reaching thousands.
If you're currently doing outreach with templates or basic mail merge, the jump to AI personalisation is easier than you might think. The key ingredients are: a clean prospect list, a clear value proposition, and a system that can pull in research data and generate emails that don't sound like they were written by a machine.
The research layer is the most important piece. Get that right and the writing quality follows naturally. Get it wrong and you end up with confidently written emails that reference the wrong information, which is actually worse than a generic template.
Want to see AI personalised outreach in action? Book a free audit and we'll show you what personalised emails would look like for your actual target market.