Rainforest Raises $4M To Build Scalable QA

Rainforest QA

Quality assurance testing is a problem that can quickly escalate beyond the resources of startups and often requires entire divisions at larger companies. It’s increasingly easy to build apps with more functionality, powered by more sources of data, with more situations a user might encounter.

Rainforest has raised $4 million to scale out its solution, which combines the power of an API with the insight that comes with having actual people look at your app. Product mangers (or other engineers on the product team) provide a plain-English explanation of what needs to be tested. On Rainforest’s end, actual people then go through the processes described, noting any problems that come up.

It’s an approach Rainforest came up with after doing contract QA work for its fellow batch companies in Y Combinator’s Summer 2012 class. The startup only came up with its central idea with three weeks left before demo day, but realized that the approach taken by most QA teams and third-party consultants doesn’t scale down to startups very well. Building automated tests for everything is great when you’re working on a stable base, but when you’re constantly rolling out new features and interface changes, it’s hard to justify dedicating resources to building out a test suite that needs to be rebuilt when you refactor the actual product.

Rainforest ramps up its testing capacity by constantly testing new workers on Amazon’s Mechanical Turk. It’s built a layer on top of the service that tests workers on the various skills required to run through actual tests for startups. It has to pay them to work through the practice tasks, but it results in having a large roster of actual people with the skill set to quickly jump into a new app and identify bugs.

The company isn’t completely against using automated testing, either — Rainforest CEO Fred Stevens-Smith says the company’s software records key presses and click paths made by people doing testing to create drafts of automated tasks. Since it’ll run through an app on its own virtual machine stack, the company can also give performance data based on logs from PHP- and Javascript-based apps.

Right now Rainforest is capable of faux-automating quality assurance testing on web apps running on Windows or Android. Stevens-Smith says Mac and iOS web support will land sometime in the next month, with actual native app support in testing with Enterprise-tier customers and scheduled for release in Q4 2015.

Rainforest’s seed funding came in from Storm Ventures, Rincon, Andreessen Horowitz, Morado, and Crosslink. Its clients include fellow Y Combinator alumni at Zenefits and Soylent and teams at IBM.