Cold email A/B testing: What to test first for maximum impact

multichannel_m · 2026-02-27 · 1,340 views

Been A/B testing cold emails for 2 years. Wasted a lot of time testing the wrong things first. Here is the priority order that actually moves the needle based on hundreds of tests.

Test #1: Subject lines. This has the biggest impact on reply rates. Test 2 subject lines per campaign. Run each for 200+ sends before declaring a winner. Test curiosity-based vs direct vs personalized subject lines.

Test #2: Email length. Short (under 50 words) vs medium (50-100 words). In my experience, shorter almost always wins for cold email. But test it for your ICP — some enterprise buyers prefer more context.

Test #3: CTA style. Question CTA ("Would this be relevant?") vs soft CTA ("Happy to share more if helpful") vs direct CTA ("Free for a 15-min call Thursday?"). Question CTAs typically win.

Test #4: Personalization depth. Generic (name + company) vs signal-based (reference funding, hiring, tech stack). Signal-based personalization usually lifts reply rates by 50-100% but costs more to produce.

Do NOT bother testing:

• Font styles or formatting — keep everything plain text
• Signature variations — minimal impact
• Send time within the 8-10 AM window — negligible difference
• From name variations — stick with first name + last name

Focus your testing energy on subject lines and CTAs first. These two variables account for 80% of the performance variance in cold email campaigns.

Comments (4)