I Built an AI That Roasts Cold Emails — Here's What 18,000 Drafts Taught Me

Ahsan Muhammad Amin — Thu, 28 May 2026 12:16:31 +0000

Most cold emails deserve to be ignored.

I know that sounds harsh. But after scoring 18,000+ cold email drafts through RoastMyEmail, the tool I built and launched this year, the data is pretty unambiguous. The average score on a first draft is 32 out of 100. That's not a rounding error. That's a systemic failure across industries, seniority levels, and company sizes.

And the worst part? Almost every bad email fails for the exact same reasons.

That's what this post is about: why I built the tool, what the data showed, and what I'd tell anyone writing cold email right now.

Why I built this

I've been on both sides of cold email. Sending it as a founder trying to get early customers. Receiving it as someone who runs a small product and apparently looks like a good prospect to every B2B SaaS company in existence.

The gap between what senders think they're sending and what recipients actually experience is enormous.

Senders think their email is personalised because they swapped in a company name. Recipients see “I love what you're doing at Acme Corp” and immediately know it's a blast. Senders think their value prop is clear because they spent three hours writing it. Recipients read “leverage cutting-edge AI to drive synergy and scale revenue 10x” and have learned nothing. Senders think asking for a 30-minute call is reasonable. Recipients see it as asking for a marriage proposal on a first date.

The problem isn't effort. Most senders are trying. The problem is that you cannot read your own cold email the way a stranger reads it. You know your product, your intent, your proof. The stranger only knows what's on the screen. That blind spot is where reply rates die.

I wanted to build something that closed that gap. A tool that reads your draft the way a stranger would, tells you exactly what's failing, and gives you a direction to fix it. Not a generic grammar checker. Not a “score” with no explanation. A brutally specific critique before the email reaches a real inbox.

That's RoastMyEmail.

How it works

The core loop is simple:

Paste your cold email draft
Get a 0–100 score with line-by-line callouts
See exactly which lines are failing and why
Get a rewrite direction
Fix it and send it with confidence

The AI evaluates four dimensions independently:

Subject line strength
Personalisation depth
Hook quality
Clarity + CTA friction

Each gets its own score and specific feedback. Not a vague “could be better” but a callout like:

“Your opener could go to 10,000 people without changing a word.”

Or:

“You've used three buzzwords in one sentence and communicated nothing concrete.”

The roasts are intentionally blunt. That's the point. Polite feedback on cold email is useless. A colleague who says “looks good to me” isn't reading it as a stranger. The AI doesn't pull punches because real prospects don't either.

What 18,000 drafts actually showed

After scoring tens of thousands of cold emails, the patterns are depressingly consistent. Here are the findings that surprised me most:

The average first draft scores 32/100

Not 60. Not 50. 32.

Most cold email going out into the world right now is failing before a human even decides whether to read it.

The most common opener: “I hope this email finds you well.”

It appears in roughly 1 in 4 drafts.

It signals bulk send before the recipient finishes the first sentence. It has never helped anyone get a reply. It persists entirely because senders default to it when they run out of things to say.

The second most common opener: “I wanted to reach out.”

Same problem.

It's a throat-clearing line that tells the recipient nothing about why this email exists. Cutting it entirely and starting with the actual point raises scores significantly.

Buzzword density is inversely correlated with clarity

This one is obvious in retrospect but striking in the data.

Emails that use words like:

leverage
synergy
cutting-edge
holistic
seamless

in the first three sentences almost always have value props that can't be explained in plain English.

The buzzwords aren't decoration. They're a signal that the sender hasn't figured out what they're actually offering.

The CTA is the most fixable problem

Asking for a 30-minute call in email one is the most common CTA pattern.

Replacing it with a yes/no question or a permission ask like:

“Should I send the one-pager?”

is the single change that moves scores the most on a rewrite.

It's also the easiest change to make. One line. No restructuring required.

The best-scoring emails share one characteristic

They read like one person writing to one person.

Not a campaign. Not a template. Not a pitch deck in email form.

One human who looked at another human's world and had something genuinely relevant to say about it.

Everything else, including length, format, and subject line structure, matters less than that.

The hardest thing about building this

The scoring model was the obvious challenge.

Getting an AI to evaluate “personalisation depth” and “hook quality” in a way that's consistent, specific, and actually useful, not just pattern-matching surface-level spam signals, took a lot of iteration.

But the harder problem was the feedback quality.

A tool that tells you:

“Your opener is generic.”

is marginally better than nothing.

A tool that tells you:

“Your opener — ‘I wanted to reach out because I noticed your company is growing’ — signals bulk send, contains zero specific research, and could be sent to 10,000 people without changing a word.”

is actually useful.

Getting the feedback to be that specific, consistently, is what took the most work.

The score is the hook.

The line-by-line feedback is the product.

The other thing I didn't anticipate was the leaderboard.

I added a public leaderboard of the worst-scoring emails, anonymised, mostly as a fun feature. It became one of the stickiest parts of the product.

People love seeing how bad the worst emails are.

It also creates a natural benchmark. If your email scores 32, you can see exactly where that puts you.

What I'd tell anyone writing cold email right now

1. Replace your opener before you change anything else

If your first sentence could be sent to 1,000 people without changing a word, it is not a cold email opener.

It is a blast opener.

Find one specific, verifiable thing about this recipient:

A post they published
A hire they made
A product change
A metric they shared publicly

Write that as your first sentence.

It takes five minutes of research per prospect.

It is the highest-return change you can make.

2. Your value prop should be one outcome with one proof point

Not a feature list.

Not a buzzword stack.

This is a value prop:

“We helped [Company] reduce onboarding drop-off by 34% in one quarter.”

This is not:

“We leverage cutting-edge AI to drive holistic revenue synergy.”

The test:

Can a stranger read it and know exactly what happens if they buy?

If not, rewrite it.

3. Make your ask smaller than a 30-minute call

Examples:

“Worth a 2-line reply?”
“Should I send the one-pager?”
“Is this on your radar for Q3?”

These are low-friction asks that a busy person can answer in ten seconds.

A 30-minute calendar hold from a complete stranger is a large ask that busy people decline by not replying.

Email one earns a conversation.

It does not earn a meeting.

Try it

RoastMyEmail is free to use.

Paste any cold email draft, subject line and body, and get a 0–100 score with line-by-line feedback in seconds.

👉 https://www.roastmyemail.fun

If you're building something in the outbound, sales, or email space and want to talk shop, I'm always up for it.

Find me at https://x.com/roastmyemail or drop a comment below.

And if your first draft scores above 70, genuinely impressive.

Most people need at least one rewrite to get there.

Just released a WordPress plugin: Circulux Product Video 🎬

Ahsan Muhammad Amin — Tue, 28 Apr 2026 17:01:06 +0000

WooCommerce product galleries are heavily image-focused by default, which becomes a limitation when store owners want to showcase product videos like demos, unboxings, or promotional clips directly inside the gallery.

Most existing solutions either rely on custom code or introduce heavy, feature-rich plugins that complicate the setup.

To address this, I’ve released Circulux Product Video, a lightweight WooCommerce extension that adds native video support directly into the product gallery experience.

💡 What the plugin does

Circulux Product Video allows you to seamlessly integrate videos into WooCommerce product galleries without changing the core user experience.

It supports multiple video sources and keeps everything unified inside a single gallery structure.

🎥 Supported features

🎬 Video Sources

YouTube URL embeds
WordPress Media Library videos
HTML5 video files (MP4, WebM, Ogg)

🖼️ Gallery Experience

Mixed media gallery (images + videos together)
Responsive carousel slider
Touch/swipe navigation for mobile users
Thumbnail navigation (grid or slider based on count)

⚙️ Key capabilities

🎥 Native video integration inside product gallery
🖼️ Unified image + video browsing experience
📱 Fully responsive design for all devices
⚡ Lightweight implementation (no unnecessary UI bloat)
🛒 Compatible with WooCommerce HPOS
🌍 RTL support included

🧠 Implementation approach

Instead of replacing WooCommerce’s gallery system, the plugin extends it using hooks and structured enhancements.

This approach ensures:

Better theme compatibility
Minimal performance impact
No disruption to existing WooCommerce workflows

The goal was to keep it simple, stable, and production-ready.

🔗 Plugin link

https://wordpress.org/plugins/circulux-product-video/

🤝 Feedback

Would love to hear how other developers are handling product video integration in WooCommerce:

Custom implementations?
Different plugins?
Or avoiding videos inside galleries altogether?

Any feedback, suggestions, or real-world use cases are welcome.

DEV Community: Ahsan Muhammad Amin

[Boost]