Prove A Human Sent It

In a spammed inbox, the differentiator is proof a person actually looked at them, not another AI-personalized template.

Every prospect's inbox is now full of email that reads like it was personalized, because it was, by a tool that pulled a name and a company off LinkedIn and merged it into a template. Readers have learned to spot the merge field even when it is dressed up well. The one thing that still gets through is proof that an actual person spent real time looking at this one specific recipient before hitting send. That takes longer to write. That is exactly why it still works.

What to do: Before you write a word, run the send through the Three Knows: know your audience (who exactly you are writing to and what they are dealing with this week), know your differentiation (why you and not the other cold email already in their inbox today), and know what good copy looks like in your category (so you read as sharper, not just louder). Then build the email against the Six Hows: a real sender name, a subject line under 60 characters, one hook that is true of this recipient and nobody else, one objection you kill before they raise it, one specific proof point, and one call to action a tired person can say yes to in ten seconds.

Why it works: Five of the Six Hows can be produced by a language model at scale. The personal hook cannot, because it requires a human to have actually looked at this one person, and that is the one thing a scanning reader is now checking for.

Example: Jon Buchan, running his agency Render Positive at the time, wrote a cold email late one night with a tone he half-expected to backfire, and sent it anyway to marketing directors at brands like Pepsi and Cisco. It worked well enough that he built a deliberate version of the stunt: a physical letter through the post, with a sticker of a ferret in a dress the internet had nicknamed Colin, followed days later by an email with the subject line "Sorry for the ferret in the post." That follow-up email hit 80%+ open rates, up from the 40% his original cold email had been pulling. The stunt became the origin story of his current agency, Charm Offensive, built on teaching this style of outreach. No AI tool generates an apology for a ferret. That is exactly the point.

Walk it through

Take the framework and run it on a real send instead of a hypothetical one.

1. Write down your Three Knows before you open your email client. Who: the actual title and situation of the person you are emailing, not a persona slide. Differentiation: the one sentence a competitor's rep could not honestly say about their own product. Category copy: pull three cold emails you have personally received in the same space and note what all three do wrong.

2. Draft the subject line first, and count the characters. "Sorry for the ferret in the post" runs 32 characters, well under the 60-character ceiling. Curiosity, not a keyword stuffed in for search. If your draft subject truncates on a phone lock screen, cut it.

3. Write the hook sentence and stress-test it. Read it back and ask if this sentence could be true of literally anyone else on your list. If yes, it is not a hook, it is a mail-merge field wearing a hook's clothes. Rewrite until the answer is no.

4. Name the objection out loud, then answer it in one line. "Never heard of you" gets answered with a proof point, not a rebuttal. "Too small for this" gets answered with a specific name or number, not a superlative.

5. End with a CTA a tired person can say yes to from their phone. Not "grab 30 minutes to discuss synergies." Buchan's own CTA skipped the calendar link and offered to buy the reader a drink instead, an absurd promise from a total stranger, and unmissable for exactly that reason.

The read

• The hook is the whole game. Everything else in the Six Hows can be templated. The hook is the only line that has to be written fresh for one person, and it is the only line the reader is actually checking.
• A short subject line is a commitment, not a trick. Under 60 characters forces you to lead with the interesting part instead of burying it behind a keyword-stuffed clause.
• Absurdity is a costly signal. A sticker of a ferret in a dress is expensive to fake, in time and in dignity. Costly signals are exactly what an automated pipeline skips, because automating them defeats the purpose.

Steal it

Take your best cold email template and delete every personalization token. Look at what's left. If the answer is "not much," you have a mail merge, not an email. Rebuild it around the Six Hows, and spend your actual research time on the hook line and nothing else. Everything downstream of the hook can be a strong, reusable template. The hook cannot, and that is where the ten minutes per prospect you were not previously spending should go.

Defend the same way on the receiving end. When a cold email lands with a hook that feels earned, that is a signal the sender did real work, and it is worth a real reply even if the answer is no. When it has a first name merged into a template sentence, filter it without guilt. Attention is the resource being competed for. Spend yours on the senders who spent theirs on you first, and build your own outbound so a stranger would say the same about you.

Gotchas

• A gimmick without a real offer behind it burns the relationship. Buchan's ferret worked because the pitch behind it was competent. A funny hook attached to a weak product gets you a one-time laugh and a permanent no.
• What works for one-to-one outbound does not transfer to a broadcast list. A hook this specific does not scale to ten thousand recipients. Reserve it for high-value, low-volume sends, not your whole newsletter.
• Physical mail plus a follow-up email costs real money and real time per prospect. Save the full version of this play for accounts big enough to justify the spend, not every lead in the funnel.