DG-201a · Module 3

A/B Testing for Sequences

3 min read

Most demand gen teams A/B test the wrong things. They test subject lines, email length, and send times — surface-level variables that produce marginal improvements. The variables that actually move conversion rates are messaging angle, value proposition framing, and call-to-action type. Testing whether "15-minute call" converts better than "quick chat" is rearranging deck chairs. Testing whether a pain-led opening converts better than a trigger-led opening is strategic optimization.

  1. Test One Variable at a Time Isolate a single variable per test — messaging angle, proof anchor type, CTA format, or channel sequence order. Testing multiple variables simultaneously makes it impossible to attribute results. One variable, two variants, minimum 100 prospects per variant, measured on reply rate and meeting conversion.
  2. Prioritize Strategic Variables Test messaging angle before subject line. Test value proposition framing before email length. Test CTA type before send time. Strategic variables affect whether the prospect engages with your core message. Tactical variables affect whether they open the email. Opening without engagement is a vanity metric.
  3. Run Tests to Statistical Significance A test with 50 sends per variant is not a test — it is a coin flip. Minimum viable test size is 100 prospects per variant for reply rate, 200 for meeting conversion. If you do not have the volume, extend the test period rather than accepting inconclusive results.

Do This

  • Test messaging angle, value proposition framing, and CTA type as priority variables
  • Run tests with minimum 100 prospects per variant
  • Measure on reply rate and meeting conversion, not open rate

Avoid This

  • Test subject lines and emoji usage as if they drive conversion
  • Declare a winner after 30 sends because one variant "looks better"
  • Optimize for opens instead of replies and meetings