Woodpecker's 2025 data says it clearly: personalised cold emails get 8.2% reply rates. Template emails get 2.4%. That's 3.4 times more responses from the same number of sends. For an SMB sending 200 emails a month, that's the difference between 5 replies and 16. Between an empty pipeline and a full one.
The kicker? AI-drafted personalised emails achieve 7.8% — nearly matching hand-crafted ones, but taking 30 seconds each instead of 20 minutes.
Why Templates Stopped Working
Templates worked in 2019 because most people weren't using them. When your inbox had one or two cold emails a week, a decent template stood out. Now the average B2B decision-maker receives 15-20 sales emails daily. Every one starts with "I noticed your company..." or "I hope this finds you well."
Your brain filters these out before you've finished reading the first line. It's not that templates are badly written. It's that they all look the same, and your recipients have developed pattern recognition that deletes them automatically.
Personalisation breaks through because it signals effort. When someone mentions your recent expansion into a new market, or references a specific product you launched last quarter, it's immediately clear they've done their homework. That's rare enough to be worth reading.
The One-Line Difference Between 2% and 8%
You don't need to rewrite your entire email for each prospect. Research shows one specific, relevant reference is enough to shift the reply rate dramatically. The rest of the email can follow a consistent structure.
What works:
- "I saw you recently opened a second office in Manchester — scaling logistics for two sites is tricky without the right systems."
- "Your new sustainability report mentions a target of net zero by 2030. Most companies I work with hit their first wall around Scope 3 tracking."
- "The rebrand looks great. Moving upmarket usually means your CRM needs to handle longer sales cycles — is that something you're dealing with?"
What doesn't work:
- "I noticed your company does great work." (Vague — works for literally anyone.)
- "As a fellow entrepreneur..." (Connection requests aren't personalisation.)
- "Hi {first_name}, I thought you'd find this interesting." (Everyone can see the mail merge.)
How AI Makes Personalisation Scalable
The old choice was painful: spend 20 minutes researching each prospect and writing a custom email (limiting you to 10 a day), or blast templates (getting ignored). AI prospecting tools remove that trade-off.
The process now looks like this: the AI reads each prospect's website, identifies relevant details (services they offer, markets they serve, recent changes), and drafts an email that connects those details to a problem you solve. You review and approve. The whole thing takes 30 seconds per email instead of 20 minutes.
At scale, that means 50 personalised emails a day instead of 10. With a 7.8% reply rate, that's roughly 4 conversations daily. In a working week, that's 20 new sales conversations — from one person, without burning out.
The Quality vs Quantity Myth
There's a persistent belief that you have to choose between sending fewer, better emails or more, worse ones. AI has made that a false choice. You can send high-quality personalised emails at volume because the research and drafting are automated. The human touch is in the review and the follow-up conversations — where it actually matters.
Ready to Personalise at Scale?
MiraReach reads your prospects' websites and drafts personalised outreach for each one — referencing what they actually do, not just their name. Every email requires your approval before it sends. Quality and volume, without the trade-off.
See MiraReach plans and pricing — from £19/month.
Frequently Asked Questions
What counts as "personalisation" in a cold email?
One specific reference to the recipient's business that couldn't apply to anyone else. Mentioning their company name isn't enough. Referencing a specific service, market, recent hire, or business challenge is what shifts reply rates.
How many personalised emails should I send per day?
Start with 20-30 from a warmed-up domain. Quality matters more than volume — 30 well-personalised emails will outperform 200 templates every time. Scale up as your domain reputation builds.
Do AI-personalised emails really perform as well as hand-written ones?
Nearly. The data shows AI-personalised emails achieve a 7.8% reply rate vs 8.2% for fully manual ones. That's a negligible difference — and the time saving (30 seconds vs 20 minutes per email) makes AI the rational choice for any team that needs to scale.
What if the AI gets the personalisation wrong?
That's why human review matters. No email should send without your approval. AI handles the research and drafting; you handle the quality check. In practice, most AI-drafted emails need minor tweaks at most — the underlying research is typically accurate.