A/B testing
Splitforce

Mobile Analytics Startup Splitforce Raises $150,000 Seed Round, Launches Automatic A/B Testing For iOS And Android

Next Story

Docker, Open Source Application Container Platform, Has 1.0 Coming Out Party At Dockercon14

Mobile analytics company Splitforce has launched an automatic A/B testing service called Auto-Optimization for iOS and Android that it says will dramatically increase the speed and efficiency of running user tests for startups.

The company, which currently serves over 250 app developers, also announced that it has raised $150,000 in seed funding led by SOS Ventures and angel investors including Simon Newstead, the CEO of Frenzoo; Steven Ho, an entrepreneur whose companies have been acquired by eBay and Yahoo; Jamie Lin, founder of accelerator AppWorks; and Guillaume Luccisano, founder of Socialcam.

Auto-Optimization includes support for user targeting. For an in-depth look at how it works, check out Splitforce’s explainer.

“From an A/B testing standpoint, this is really useful because if you are looking to personalize an app for different segments of users, you are running different tests on different segments, and for each segment you have to spend time managing that test,” explains founder Zac Aghion.

“But with Auto-Optimization, you can actually automate the test for all your user segments without having to actively manage them.”

Auto-Optimization is based on a class of algorithms called Bandit algorithms, Aghion explains, and it emerged from the multi-armed Bandit problem.

“It’s an English term for a slot machine and the problem is basically that you have a multi-armed slot machine. Imagine one with many different levers you can pull and each lever has a different rate of payout. One lever might pay out 25% or 10%, and you don’t know what the different payouts are. The problem is really how do you maximize the total payout by pulling the levers in some particular way,” says Aghion.

He adds that the algorithm Auto-Optimization uses becomes more precise over time using sampling.

Some of Splitforce’s competitors include Optimizely, Taplytics and Apptimize, all of which have been covered by TechCrunch.

While most major companies that offer A/B testing software have started moving toward a Bandit algorithm approach, Aghion says Splitforce differentiates with Auto-Optimization, as well as offering support for both iOS and Android native apps and Unity, the popular game engine.

Next on Splitforce’s list is support for Swift, Apple’s new iOS programming language. The startup has already built support, but won’t be able to release it until this fall because of Apple’s terms and conditions. The company is adding support for mobile web and hybrid apps based on a JavaScript library next.

Image by PicJumbo