Subject Line Lab

Fifteen cold email subject lines for one offer, grouped into five angles, each with the reason it works and the risk it carries — plus the two-line A/B test to run first and what the result would tell you.

A Markdown subject-line set containing: (1) exactly fifteen subject lines of six words or fewer, grouped three each under the five angle headings Problem, Mechanism, Number, Curiosity and Peer; (2) a why-it-works and a how-it-fails clause under every line; (3) a recommended first A/B pair drawn from two different angles with the decision each outcome drives; (4) a spam-risk note.

# Subject lines — SOC 2 evidence automation

## Problem
1. **four weeks of screenshots** — names the exact ritual the buyer dreads; fails if they have already automated it.
2. **audit window again** — the word 'again' does the work; fails on a first-time audit.
3. **evidence collection season** — insider framing, reads as peer; fails if the reader is not the one doing collection.

## Mechanism
4. **evidence pulled, not gathered** — states the product's whole idea in four words; fails if 'evidence' is ambiguous outside compliance.
5. **your cloud already knows** — implies the work is redundant; fails by being slightly too clever to parse at a glance.
6. **screenshots to API calls** — concrete before-and-after; fails with a non-technical buyer.

## Number
7. **second audit, half the time** — pairs a milestone with a claim; fails if you cannot substantiate the half.
8. **[PROOF POINT NEEDED] hours per audit** — placeholder deliberately left; a real figure here is the strongest line in the set.
9. **90 controls, one export** — specificity signals you have done this; fails if their framework count differs.

## Curiosity
10. **the pre-audit scramble** — universal and low-risk; fails by being slightly generic.
11. **what your auditor accepts** — hints at insider knowledge; fails if it reads as a threat.
12. **before the window opens** — timing pressure without a deadline; fails outside audit season.

## Peer
13. **how other infosec teams** — peer framing, deliberately unfinished; fails if it reads as a mail-merge.
14. **second-audit playbook** — offers an artefact, not a meeting; fails if you do not actually have one.
15. **your peers stopped screenshotting** — mild social pressure; fails if it reads smug.

## Run this A/B first
**Line 1 (Problem) against line 4 (Mechanism).**

- If Problem wins, your market buys on pain and the body copy should stay on the four weeks before it ever mentions how the product works.
- If Mechanism wins, your buyers already accept the pain as a fact of life and are shopping for a specific approach — move the technical detail up and lead the body with the integration list.

This pair is worth running before any body-copy test, because it settles which of the two arguments the entire sequence should be built on.

## Spam risk
Nothing here uses a classic trigger term. Watch line 8 once the number is filled in: a digit plus a percent sign in a subject line is the most common filter trip in this set, so write it as words if the figure is a percentage. Avoid adding 'free', 'guaranteed' or an exclamation mark to any of these lines.