I’m Kayla, and I run emails for two small shops and a little course on the side. I live in my inbox. So I test a lot. ActiveCampaign’s split testing helped me stop guessing and start sending stuff that people actually click. (If you’d like to see the nuts-and-bolts breakdown, here’s my full ActiveCampaign split-testing deep dive.)
You know what? It felt a bit nerdy at first. But once I saw the numbers, I was hooked.
The short sweet version
- It’s easy to test subject lines, send times, from names, and email content.
- Automation splits helped me see which path sold more.
- I got real gains fast, but setup can feel fussy the first week.
Now let me show you what I did, with real examples.
If you're looking to level up every part of your conversion flow, the tutorials over at Optimization World are a goldmine.
1) Subject lines: emoji vs plain
For my winter gear sale (12,486 people), I tested three subject lines:
- “⛄ Last chance: 24-hour Winter Gear Sale”
- “Winter gear sale ends tonight”
- “Sale ends tonight: winter gear”
Winner after 4 hours: the snowman line.
- Open rate: 38% (emoji) vs 29% and 27% (plain)
- Click rate: 5.2% (emoji) vs 3.9% and 3.6%
- Sales (Shopify synced): $6,240 vs $4,010 and $3,650
Why it worked: short, clear, a tiny bit fun. My crowd likes cozy. I do too.
Small note: I also changed the preheader. ActiveCampaign doesn’t test that as a separate field for campaigns, so I just cloned the version and wrote a new preheader. Not hard. Just a little messy.
2) From name test: person vs brand
I ran this on my candle shop list (8,302 people).
- “Kayla at Meadow Wick”
- “Meadow Wick News”
Winner: “Kayla at Meadow Wick.”
- Opens: 33% vs 26%
- Clicks: 3.4% vs 2.5%
It felt more human. Also, my mom said it looked like a note from me, not a flyer. Moms are usually right.
3) Send time: morning vs night
I sell to two types: teachers and parents. I split on local time.
- 8:00 AM vs 8:00 PM
For teachers (school supplies list):
- 8 AM won. 41% opens vs 27%. Clicks 4.9% vs 3.1%.
For parents (home goods list):
- 8 PM won. 35% opens vs 30%. Clicks 3.6% vs 3.2%.
ActiveCampaign also has Predictive Sending (I’m on the Professional plan). It helped a bit on the parent list, but plain 8 PM still beat it by a hair that week. Funny, right? I still use Predictive Sending when I’m short on time.
4) Button copy: “Get My Code” vs “Shop Now”
Same email, two buttons. I only changed the words.
- “Get My 10% Code”
- “Shop Now”
Winner: “Get My 10% Code.”
- Click rate: 4.8% vs 3.1%
- Sales: $3,220 vs $2,140
People like getting something. Shocking, I know. I made the button green for both versions. No trick colors.
5) Automation split: SMS nudge vs extra email
In my 6-email welcome series, I added a split after Email 2:
- Path A: send a short SMS at 6 PM (“Hey, it’s Kayla—your free wick trimmer is still in your cart”)
- Path B: send a short FAQ email instead
Winner: SMS (I pay per text, so I checked costs).
- First purchase rate in 7 days: 7.8% (SMS) vs 5.1% (FAQ)
- Added cost for SMS: $47
- Extra revenue: about $1,180
- Worth it? Yep.
Note: setting SMS needs numbers and consent. I had that. If you don’t, use a short reminder email with a big button. Still works.
ActiveCampaign even has an AI-powered split tester for automations that can pick the winning path automatically if you’d rather let the robots crunch the numbers for you.
6) Abandoned cart offer: free ship vs 10% off
I worried about margin. So I split test inside the automation. I triggered each path with the Send an email action—their official split-testing guide walks you through the clicks if you need a refresher.
- Path A: free shipping code
- Path B: 10% off code
Winner: free shipping.
- Recovery rate: 12.4% (free ship) vs 11.2% (10% off)
- Margin saved per order was better with free ship on items under $40.
- On orders over $100, 10% off did better in revenue. So I kept both and used a rule by cart value. Felt fancy, but it paid off.
If you’re curious how straight-up price experiments can play out, here’s a candid write-up on split-testing product prices.
7) Black Friday timing: 3-hour winner test
For my big sale, I sent 20% of the list first, split two subject lines, let it run for 3 hours, then ActiveCampaign pushed the winner to the rest.
- “Black Friday: 30% off sitewide + free gift”
- “30% off + free gift (today only)”
Winner: the second one.
- Opens on sample: 44% vs 39%
- Clicks on sample: 7.9% vs 6.2%
- Full send kept the lead and finished strong
Pro tip: set the winner by clicks, not opens. Opens can be weird now with privacy stuff. Same idea works great on landing pages too—see this ClickFunnels page split-test for inspiration.
What bugged me a bit
- The split block in automations is strong, but it won’t “auto-pick a winner” and switch the whole flow by itself. I had to check stats, then change the paths.
- Reports can load slow if you filter by device and tag and date. I got a coffee. It helped me and the report.
- Testing preheader alone takes a clone. I wish it had a separate field in the A/B tool.
- Naming. Please name every test like “BF-2024-Subject-Emoji vs Plain.” I learned that the hard way when I built the same test twice.
Tiny things that helped a lot
- Keep a control. Change one thing at a time if you can.
- Let tests run long enough. I try 4–24 hours, depending on list size. I learned that the hard way while running onboarding experiments in Mixpanel.
- Pick the right winner metric. I use clicks for campaigns, purchases for automations.
- Use segments. Parents vs teachers, new vs repeat. Different timing, different wins. For a totally different vertical, imagine curating messages for a discreet dating audience in Kent—this sugar daddy guide for Kent shows how laser-focused, location-based insights can elevate outreach and engagement, and you can borrow the same hyper-local mindset for crafting email segments that convert.
- Save winning parts as blocks. ActiveCampaign makes that easy, and it saves time.
Real results after 6 weeks
- Average open rate: up from 26% to 33%
- Average click rate: up from 2.8% to 4.1%
- Welcome series sales (first 7 days): up 31%
- Abandoned cart recovery: up from 9% to 12%
Not perfect. But very real.
Should you try it?
If you send more than one email a month, yes. Even one test per send can help. Start simple: subject lines and send time. Then try a split in your welcome or cart flow.
Here’s the thing: split testing won’t fix a weak offer or a dull product shot. But it will help a good message get seen, and a clear button get clicked. And that matters.
I still mess up names now and then. I still ship with typos sometimes. Mistakes are human—whether you’re pressing “send” on a campaign or figuring out first-time chemistry with someone new—and a quick reality check can be both helpful and hilarious. For a tongue-in-cheek look at slip-ups in a very different