A/B Testing
for React
Make informed design decisions based on real interaction data. Minimal set up, maximum insight.
We'll split your app's users into two groups and render one of your A/B test variants to them. So...
<ABTest
abTestId="abt_e91f9f007bd2096875ebe6ec2943fd6d"
variantAId="abv_e0bad4b1f1fb69d0a531359043d9fe1a"
variantBId="abv_b9b5cabcf7ccabe12301bce31e270b84"
>
<button className="light-blue">Add to cart</button>
<button className="deep-blue">Add to bag</button>
</ABTest>
Test components, not pages
A/B tests are usually run by creating two variations of a page and splitting visitor traffic between them. That's fine if you're testing landing pages, but when you're testing the way users interact with your site or app, you can't always send them to a different URL.
Not all clicks are conversions
A single function call is all it takes to create a conversion, allowing you to convert based on events like successful API calls, time on screen, and anything else you can dream up.
Persisted rendering
We'll make sure that your users see a consistent UI. Once they've been presented with a variant of your component, they'll see it forever.
No extra props
Your components will work exactly as you code them. Preferr simply wraps the components and listens for conversion events in the background.
Keep it simple
You don't need to be a statistician or marketing maverick to get useful information out of an A/B test.
Numbers you want
Get meaningful stats like conversion lift, conversion rates, and a p-value calculation so you know how statistically significant your data is.
What learning curve?
There are no lengthy tutorials, excessively featured user interfaces, or confusing terminology. Documentation is short and sweet and you'll be up and running in minutes.
On your terms
Enrich your tests by optionally identifying visitors with your own data and terminology.
- Total Visitors
- 2,304
- Conversion Lift
- 25.8%
- p-value
- 0.0327
See your users preferences
Preferr keeps track of what your rendered variants look like so that you can quickly see what's working and what isn't.
