Most cold email campaigns launch with one version and never improve. The teams that consistently hit 5-10% reply rates test everything — subject lines, opening lines, CTAs, and timing.
Here's how to run meaningful A/B tests on cold email.
Why A/B Testing Matters for Cold Email
Small improvements compound. A subject line change that lifts reply rate from 3% to 5% on a 1,000-contact list means 20 extra replies — which could be 5-10 extra meetings.
What to Test (In Priority Order)
1. Subject Lines
The highest-impact variable. If they don't open, nothing else matters.
Test: short vs. long, question vs. statement, personalized vs. generic.
Example A: "quick question about {{company}}"
Example B: "{{firstName}}, idea for your sales team"
2. Opening Lines
The first sentence determines whether they keep reading.
Test: observation-based vs. pain-point-based vs. compliment-based.
Example A: "Noticed {{company}} is hiring 3 new SDRs..."
Example B: "Most teams at your stage struggle with reply rates dropping as volume goes up."
3. Call to Action
The ask determines whether they respond.
Test: soft ask vs. hard ask vs. specific time proposal.
Example A: "Worth a quick chat?"
Example B: "Would Tuesday at 2 PM work for a 15-minute call?"
4. Send Timing
When you send affects when (and if) they see it.
Test: Tuesday 9 AM vs. Thursday 2 PM. Morning vs. afternoon. Weekday vs. Sunday evening.
5. Sequence Length
More steps means more touchpoints — but also more chances to annoy.
Test: 3-step vs. 5-step sequences on similar audiences.
How to Run a Proper Test
- Split your list evenly: 50/50 random split. Don't test on different audiences.
- Change one variable: If you change subject AND opening line, you won't know which one mattered.
- Use enough volume: At least 100 contacts per variant. Below that, results aren't statistically meaningful.
- Measure reply rate, not open rate: Opens are unreliable. Replies are the metric that matters.
- Run for the full sequence: Don't judge after step 1. Most replies come on steps 2-4.
A/B Testing in ClickReach
ClickReach doesn't have a built-in A/B testing button, but the manual approach works well:
- Create two sequences with identical contacts split 50/50
- Change one variable (e.g., subject line) between them
- Run both for the full sequence duration
- Compare reply rates in the per-sequence analytics view
- Kill the loser, scale the winner
This manual method gives you full control and lets you test anything — not just what the tool's A/B feature allows.
