I tested 14 cold email subject line variants across 8,400 sends to SaaS buyers and one pattern won every single week for 3 months without a meaningful challenger
subject_petra · 2026-09-02 · 0 views
I've been systematically A/B testing cold email subject lines for three years. Most tests produce small, noisy differences. Subject lines that look clever in theory perform the same as boring controls. This test was different.
I sent 8,400 cold emails over 12 weeks to VP-level buyers at SaaS companies with 20 to 200 employees. Fourteen subject line variants tested in rotating pairs against each other. Here's what I found.
The winner, every week for all 12 weeks: the prospect's first name only. Nothing else. Lowercase. No punctuation.
"sarah" as a subject line consistently outperformed every other variant I tested. By a margin that increased over time, not decreased. Second place was their company name in lowercase. The winning margin over second place held at around 0.4 percentage points in reply rate, every single week.
What the first-name-only subject line does: it reads as a message from someone who knows you. It looks like the kind of email you get from a colleague or someone you met at an event last month. It doesn't trigger the pattern-matching the brain runs on marketing email subject lines, because it doesn't look like one.
What I tested against it and lost: "quick question," "your [title] stack," "re: [relevant topic]," "idea for [company]," and nine other variations that are popular in cold email communities. All of them performed worse. Several performed significantly worse on this ICP.
This result replicates. I've since used the first-name-only subject line across six client campaigns targeting different ICPs. It wins or ties every time. The only ICP where it didn't clearly win was professional services, where the firm name in lowercase performed equally well.
One implementation note: make sure your personalization variable pulls cleanly. A broken first_name variable rendering as literal text in your subject line is a campaign killer. Test your sequence in Instantly or Smartlead with a seed inbox before going live. This matters more when the entire subject line is nothing but the variable.
I run these tests from 24 PuzzleInbox Google Workspace inboxes. Infrastructure stability is a prerequisite for clean subject line testing. If your deliverability is fluctuating because of inbox problems, your test results are noise, not signal. Get the infrastructure right first, then run the tests.