Join the waitlist
Menu
Free tool

Grade a cold email against Dylan's rules.

Paste a first email. The grader checks it against the nine rules Dylan writes by and marks every line it finds. It runs in this browser: what you paste is not sent anywhere.

Stays on this device

Your first email
Who sends it?

0 words · 0 sentences

Your score

Your score and the fixes show here.

The rubric

The nine rules it checks.

Each rule is one the site already states, quoted from where it says it. The weights add up to 100.

  1. Rule 1: Who is writing, and for whom, in the first line

    20 points

    “Every first email Dylan sends says, in its opening line, that he is an AI agent and which company he works for.”

    From an AI agent: the first sentence says it is an AI and names the company. From a person: your name and your company in the first two sentences.

    Why:How Dylan writes: the voice rulesAI disclosure in cold email

  2. Rule 2: A reason to write to this person, now

    20 points

    “…a reason that belongs to this person, this month: a new site, a new leader, a new market.”

    A reason (because, noticed, just opened) next to a detail about this reader in the first three sentences. A line that could go to anyone costs half; an empty merge field costs all of it.

    Why:What survived the AI SDR hype

  3. Rule 3: One small ask

    15 points

    “One small ask”

    Exactly one question or request gets full marks, two get half. Dylan's example first email asks one question. The way-out line never counts as an ask.

    Why:A first email, line by line

  4. Rule 4: An obvious way to say no

    15 points

    “Short, plain sentences, a specific reason to write, and an obvious way out.”

    A line like: If not, reply "no" and I won't write again. An unsubscribe or opt out line counts too.

    Why:A first email, line by line

  5. Rule 5: Short enough to judge at a glance

    5 points

    “Five sentences carry four things: who is writing, for whom, why this person now, and how to say no. A stranger can judge it at a glance.”

    Up to 90 words gets full marks, 91 to 150 half, more none. A sentence over 30 words halves it. The greeting line and the signature are not counted.

    Why:A first email, line by line

  6. Rule 6: Plain: no attachments, few links, no images

    10 points

    “Write plain emails: no attachments, few links, no tracking pixels.”

    0 or 1 link gets full marks, 2 half, more none. An attachment, an image or a tracking pixel: none. Images and hidden links are seen only when you paste from an email app.

    Why:Cold email deliverability in 2026Powered by Folderly

  7. Rule 7: A subject line a person would type

    5 points

    “Subject lines are lowercase. The way a person types them. No fake "Re:" pretending to continue a conversation.”

    Lowercase gets full marks, one or two capitalized words half. Re:, Fwd:, an exclamation mark or Title Case: none. Short acronyms like AI or CRM are fine. No subject: not graded.

    Why:How Dylan writes: the voice rules

  8. Rule 8: Plain words, no hype, no guarantees

    5 points

    “When a sentence sells instead of describes, cut it.”

    The words the site cuts (revolutionary, seamless, AI-powered and the rest) and claims like "guaranteed" or "risk-free". One costs half; two or more cost all of it.

    Why:The design page: voice

  9. Rule 9: No fake follow-up

    5 points

    “…no "just following up on my last note" when there was no last note.”

    Phrases like "just following up", "circling back" or "bumping this" in a first email. One also holds the score at 59.

    Why:How Dylan writes: the voice rules

The weights and the length lines are this grader's own, set from Dylan's example first email (63 words, five sentences). They are not a benchmark.

What this grader cannot see

  • Whose domain sends, warm-up, sending caps, send windows, SPF, DKIM and DMARC. They are not in the text of an email. Check a domain's records
  • Whether a name is real. Dylan's rule is never a fake name, never a fake person. Text cannot show it.
  • The From line. You paste the body, not the headers.
  • First person and American spelling. That is Dylan's own voice, not a rule for every writer.
  • Whether the reason is true, and whether the offer fits. Only the sender can know.
  • The law. The grader gives no legal verdict on CAN-SPAM or the EU AI Act. The post on AI disclosure covers both, and it is not legal advice either. AI disclosure in cold email

What it never claims

  • No reply, open or meeting forecast, and no inbox test. The score says how closely an email follows Dylan's rules, nothing else.
  • No comparison with other emails. There is no data behind one.
  • No AI detection. Who sends it is your own answer; the grader never guesses who wrote the text.

Start from a template

Cold email templates written by the same rules, each with an example you can grade here.

Want Dylan to write the first email for you?

Tell Dylan your website and who you want to meet. Your plan shows the first email he would send, and nothing is sent without your approval.

Already know your definition? Send the full brief