When Google Optimize shut down in September 2023, a lot of small businesses lost the only A/B testing tool they could afford. It was free, it was good enough, and it worked.
Then it was gone.
The assumption was that people would move to one of the paid alternatives. And some did. But if you actually went looking for what was available, the options were bleak. VWO starts around $200 a month. AB Tasty is in a similar range. Optimizely doesn't even list pricing on their site, which usually means you can't afford it.
For a company with a marketing team and a testing budget, $200 a month is nothing. For a freelancer running a portfolio site, or a small business with one landing page and a hunch about whether a different headline might work better? That's a lot of money to find out.
The gap
There's a weird hole in the market.
On one end, you've got enterprise tools with dashboards full of segments, audiences, and multi-armed bandits. On the other end, you've got nothing. Free tools basically disappeared when Google pulled the plug.
The people in the middle, the ones who just want to test a headline or a button or a signup form, got left behind. Not because the technology is hard. A/B testing is a pretty simple concept: show half your visitors version A, show the other half version B, count who clicks. The tools just aren't built for people with small sites and small budgets.
What we wanted to build
We wanted something that a freelancer could set up in five minutes. One script tag, a short setup wizard, and a clear result page that doesn't require a statistics degree to read.
The scope is intentionally small. SplitPea does one thing: it helps you run website experiments and tells you which version won.
There's no heatmaps, no session recordings, no audience segmentation, no integrations page with forty logos on it. Those are fine features. They're just not what most small businesses need when they're trying to figure out if their pricing page headline is any good.
We also wanted it to be cheap. Actually cheap, not "affordable" in the way enterprise software uses that word.
The free plan lets you run one experiment on one site. Paid plans start at $29. The most expensive plan is $49. We think that's about right for what most people need.
How it works (briefly)
You add a small JavaScript snippet to your site. It's under 8KB and loads asynchronously, so it won't slow anything down.
Then you create an experiment:
- Give it a name
- Write a hypothesis if you want to (it's worth doing, even if it's just one sentence)
- Pick the page you want to test
- Set a conversion goal
- You can change elements visually by clicking on them, or use a CSS selector if you prefer.
SplitPea splits your traffic, tracks conversions, and uses Bayesian statistics to figure out when one version is reliably beating the other. When confidence is high enough, it tells you. You decide what to do with the result.
That's the whole thing.
Who it's for
Freelancers testing headlines on their portfolio sites. Small businesses trying to figure out which version of their booking page gets more enquiries. Agencies that need to show clients what's working. Anyone who's ever changed something on their website and wondered if it actually helped.
If you've never run an A/B test before, SplitPea is a good place to start. If you used to use Google Optimize and haven't found a replacement, this is what we built it for.
What we believe
A/B testing should be accessible to people who don't have analytics teams. You shouldn't need to spend $200 a month to find out if a different button colour gets more clicks. And the tool you use to do it shouldn't be more complicated than the test you're running.
We built SplitPea because that gap in the market seemed wrong. It still does.