Generic Cold Email: The Two Swap Test

What you'll get

1 copy-ready blocks
3 min read

Two swap test

How to use it

The hack

before a first email goes to a named account, have your AI assistant run two swaps: the buyer's company for one of its competitors, then your company for one of yours. Every line that still reads fine after both swaps was written for nobody. Rebuild the email around the line that broke.

The hack: before a first email goes to a named account, have your AI assistant run two swaps: the buyer's company for one of its competitors, then your company for one of yours. Every line that still reads fine after both swaps was written for nobody. Rebuild the email around the line that broke.

Why two swaps

A first email to a named account makes two claims: this is about you, and it could only come from us. Each swap tests one claim. If the message survives with a competitor's name in place of the buyer's, your research never made it into the text. If it survives with a rival's name in place of yours, the email says nothing a rival could not say. Most drafts survive both, which is why most first emails read like everyone else's.

The prompt

Two swap test

Below is a draft first email from [my company] to [account].

Swap 1: replace every mention of [account] with [a named competitor of theirs]. List each sentence that is still accurate and sensible after the swap.

Swap 2: in the original draft, replace [my company] with [a named competitor of ours]. List each sentence a reader would still accept after the swap.

Then tell me:
1. Which sentences survived both swaps.
2. Which single sentence broke hardest, and why.

Do not rewrite the email. Grade it only.

Draft:
[paste]

Use real competitor names in the brackets. The assistant only needs them to run the swaps.

Read the result

The sentences that survived swap 1 are not about the account: the opener about their industry, the line about companies like them. The sentences that survived swap 2 are not about you: the credentials paragraph, the claim every vendor makes. Cut the survivors. The email gets shorter and better at the same time.

The sentence that broke hardest is your message. It is usually the one that quotes something true and dated about that company and connects it to the specific thing you would do about it. Rebuild the email as that sentence plus one question, and send that.

If every sentence survives both swaps, you do not have a writing problem. You have a research problem: nothing in the draft came from looking at this company. Go back to the account's news, filings and your own notes until there is a reason to write, then draft again. The trigger event list names the 12 reasons worth watching for.

Three mistakes to avoid

Fixing a survivor by inserting the company name. "Firms like [account] struggle with reporting" survives the swap with the name changed, because the name is a label, not a fact. The fact has to change, not the greeting.
Running it on a message meant for 200 companies. A template fails this test by design. The test is for first emails to accounts that matter, where a thin first impression costs more than the hour of research would have.
Letting the assistant rewrite the draft. Asked to fix the email, the model fills the gaps with plausible lines about efficiency and growth, which are exactly the sentences that survive swaps. It grades; you write. The strip the AI tells hack covers what happens when it writes.

Make it a habit: no first email to a Tier 1 account goes out until both swaps have run, and the survivors are cut rather than defended.

Why this works on named accounts

On a lead list a generic email is not a mistake, it is the design: one message, a thousand names, and the arithmetic does the rest. On a named account the same email is a spent chance. You chose this company, which means you will want to write again, and the first message sets what every later one is worth. The swap test exists because generic is invisible from the inside: every draft feels specific to the person who wrote it. Swapping the names is how you see your own email the way the account will.

How account-based selling works →

More from the Academy

Sales hack3 min read

Strip the AI Tells Before You Send

For sellers who draft with AI and suspect the buyer can tell.

You leave with seven tells, one prompt that flags them in your draft and a rewrite rule that keeps your voice.

1 copy blockOpen and copy →

20-minute trigger sweep (every Monday)

1. Open the Tier 1 list. For each account, check: news page, LinkedIn company page, careers page.   (10 min)
2. Write down every change with its date and link. No link, no trigger.                             (5 min)
3. For each trigger: who does it land on, and do we know them?                                      (3 min)
4. Pick the three accounts with the freshest trigger. They get this week's research hours.          (2 min)

Sales hack3 min read

The Trigger-Event List: When to Approach an Account

For sellers with a good account list and no idea which account to work this week.

You leave with 12 trigger events, where to spot them and who to approach when one lands.

2 copy blocks · Spreadsheet-readyOpen and copy →

Want this in your inbox?

Optional. We use your email only to send you this, nothing else.

The Two Swap Test: Catch a Generic Cold Email Before a Named Account Sees It