Splitsense
AI that turns traffic into more revenue while you sleep
Start new thread
George Field

Splitsense - AI that turns traffic into more revenue while you sleep

by
Seio
•
Connect your website and let AI automatically generate, test, and optimise variations, perform structured A/B & optimisation tests, all without writing any code. Increase conversions without writing a single line of copy.

Add a comment

Replies

Best
George Field
Maker
📌
Hey Product Hunt! 👋 I'm George, one of the co-founders of SplitSense. I was already doing manual testing and trying different copy variations across DevRemote and a few other sites I was running, trying to figure out what actually helped my website convert traffic. It worked, but it was slow, messy, and completely unscalable. So I built SplitSense to automate exactly that. Drop in your URL, and the AI analyses your site, picks the highest-impact pages to test, then generates copy variants automatically. No dev, no complicated setup, just experiments running in the background while you get on with everything else. Would love to hear from anyone who's been through the same frustration, manually tweaking headlines, guessing at CTAs, never quite sure if it made a difference. What's the biggest thing that's stopped you from testing your copy properly? Happy to answer anything. 🙏
Becky Gaskell

This sounds interesting. I always wonder with tools like this, how much control do you actually have over what the AI is testing and changing?

George Field
Maker

@becky_gaskell Great question. The idea behind this is that its completely autonomous, so you don't have much control, but this is by design. We analyse your websites data under the hood to figure out what to test and then tell you why. However, its of course something we can look at

Becky Gaskell

@george_field That makes sense. I can see how that would be useful, but I’d probably still feel a bit nervous giving up that much control at first.

George Field
Maker

@becky_gaskell Its a good point. Something we should add into our roadmap is around controls encase users want more control over what is happening at first. We do however give the user the option to stop the experiment from running and they can see what is going to change before an experiment starts running.

So, currently, Splitsense will use your web analytics data to see what it thinks it should test, then it will go and create an experiment, generate variations, then show you what it has come up with. The different variations won't be live though unless you approve it.

Sam Sutton
Maker

@becky_gaskell Hey Becky! Adding to George's reply, our intent is to enable everyone to test & change copy to find what works best for their product regardless of familiarity with, or skill level in, copy writing and data analysis. That's why the current experience is hands-free until the key review & approval stage before anything changes on your site.

We're definitely open to providing more fine-grained control for power users though!

Becky Gaskell

@sam_sutton_ That actually sounds like a really good balance. Having that approval step would make a big difference, I think that’s where a lot of the hesitation comes from.

Maxwell Timothy

Love this! Does it work on any type of website?

George Field
Maker

@maxwell_timothy Hey Maxwell, thanks for taking the time to reply, it does indeed, you simply get the code snippet after onboarding and put that onto your website and thats it

Jared Salois

Curious whether you can actually push back on what the AI decided to test - not just approve or reject the variant, but understand the reasoning behind why it picked that specific thing to change in the first place?

George Field
Maker

@jared_salois Hey Jared, firstly, thank you for the feedback. So, at first the initial variants are more guesswork based on other experiments within the platform and a specific agent we have setup for creating an initial experiment.

However, as we gain more data from both experiments on your website and data about your traffic, users, sources ect, we start to leverage that to inform the decisions behind the experiment and we do create reasons for why a variant was created but we don't yet display it.

Jared Salois

@george_field That makes sense. Showing reasoning that’s still partly guesswork would probably create more confusion than trust. Better to surface it once the data behind it is solid enough to actually explain what’s going on.