---
title: "A/B testing opening lines | ColdDart"
description: "Compare a research-based and a challenge-based opening line on your own outcomes. A winner is named only with 30 delivered emails per variant and a significant difference."
image: "https://colddart.com/og-image.png"
---

# Compare two kinds of opening line on your own list

Variant A refers to the person's role or company. Variant B opens with a challenge. You decide with your own outcomes.

[Talk to the founder on WhatsApp](https://wa.me/5548992091242?text=Hi!%20I%27d%20like%20to%20talk%20about%20ColdDart%20%28personalizing%20the%20opening%20lines%20of%20my%20cold%20email%20list%29.%20I%20came%20from%20this%20section%3A%20A%2FB%20testing)

## Two prompt styles

ColdDart can write each lead's line in one of two styles. Variant A is research-based: it refers to something specific about the person's role or company. Variant B is challenge-based: it opens with a claim about their industry or role that is meant to make them think.

## Measuring with your own results

After you send, import an outcomes file with two columns: `email` and `outcome`. The outcome is one of `no_reply`, `replied`, `positive_reply` or `bounced`. You can also set outcomes lead by lead. ColdDart then shows the reply rate per variant.

It names a winner only when each variant has at least 30 delivered emails and the difference passes a two-proportion z-test (|z| of 1.96 or more). Until then it says there is no significant difference yet. Expect it to say that often on small lists.

## What this is and is not

It compares two ways of writing the opening line on your own data. It does not tell you what reply rate to expect, and ColdDart makes no such promise.