You're sending cold emails and getting some replies, but you know you could be doing better. You've heard about A/B testing subject lines. You know it matters. But when you sit down to actually do it, you get stuck - what exactly are you testing? How many emails do you need to send? How do you know if the results are real or just luck?
This is the frustration I see with most people trying to optimize their cold email. They either don't test at all, or they test in a way that wastes time and leads them to the wrong conclusions.
Here's what you actually need to know to test subject lines properly.
Your subject line is the only thing that determines whether someone opens your email or deletes it. Everything else - your copy, your offer, your positioning - none of that matters if the email never gets opened.
Most people know this. What they don't know is how to test it without wasting weeks and getting meaningless data.
The biggest mistake I see is testing too many variables at once. Someone will change the subject line AND rewrite their opening line AND adjust their CTA, then wonder why they can't figure out what actually worked. That's not testing - that's chaos.
The second mistake is not sending enough volume. If you send 50 emails with subject line A and 50 with subject line B, the difference between 16 opens and 18 opens doesn't mean anything. It's noise.
Let's fix both of these problems.
This is non-negotiable. You're testing ONE thing. Just the subject line. Everything else - the email body, the sender name, the send time - stays exactly the same.
If you change multiple things, you'll never know which one caused the difference in opens. You'll end up making decisions based on guesses.
Don't test a good subject line against a bad one. Test two approaches that you actually think could work.
Here are some common angles to test against each other:
Pick two that feel genuinely different. You're not testing tiny tweaks - you're testing real strategic differences.
Split your prospect list into two equal groups. Make sure it's random. Don't put all your "hot" prospects in one group or all companies of a certain size in another.
If you're using a good email platform, this is built in. Most will let you create variants and split your send automatically.
This is important. Send at least 200-300 emails with each subject line. Ideally more.
Why? Because open rates typically range from 20-40%, and you need enough data points to see a real pattern. If you only send 50 emails per variant, a single person opening your email changes the results by 2%. That's meaningless noise.
Send 300 emails per variant, and you're looking at 60-120 opens per group. Now a difference of 5-10 opens actually tells you something.
Let both variants run for at least 3-5 days before you draw any conclusions. Most opens happen within the first 24 hours, but you want to see the full picture.
And please - don't check your results every two hours. You'll drive yourself crazy and probably make a bad decision based on incomplete data. Set a calendar reminder to check after a week.
After a week, you'll have data. Let's say subject line A had a 28% open rate and subject line B had a 35% open rate. Is that a real difference, or did subject line B just get lucky?
Here's the simple rule: if the difference is less than 3-5 percentage points, treat it as a tie. The smaller your sample size, the more cautious you should be.
If you sent 300 emails per variant and you see a 7+ point difference, that's real. That's the subject line to use going forward.
If you see a 2-3 point difference, it might be real or it might be luck. Do another test. Send it again with the same variants to a different group of prospects and see if the pattern holds.
Once you've found a subject line that wins, use it. Keep using it with new prospect lists.
But here's the thing - subject line performance varies by industry, company size, and even the specific prospect list you're reaching out to. A subject line that crushes it with tech companies might underperform with law firms.
So test the winner against new alternatives as you continue sending. You're not doing this once and forgetting about it. You're continuously trying to improve.
A 5-10 point improvement in open rate doesn't sound huge until you do the math.
If you're sending 500 cold emails a month at a 25% open rate, you're getting 125 opens. If you improve to a 35% open rate through better subject lines, that's 175 opens. Fifty more conversations from the same amount of outreach.
That directly translates to more replies, more meetings, and more clients.
The catch is that testing takes time. You have to send volume, wait for data, analyze results, and iterate. If you're running sales for your own business, that's time you could be spending on other things.
That's why a lot of service businesses and agencies decide to outsource this entire process - the testing, the optimization, the continuous improvement. You focus on closing deals while someone else handles the infrastructure, lead sourcing, copywriting, and campaign management to get consistent cold email results month after month.
Ready to Sign Clients On-Demand?
BEC Growth builds and manages your entire cold email system from infrastructure to reply handling.
Book a Call →