'Benefiting from a culture of experimentation Patti Chan, VP Product at Imperfect Foods, wanted to take an experiment-driven approach to product development. Originally, they tried to create their own feature flagging system but weren’t getting as much value out of it as they hoped. “Split allowed us to establish an “experiment-driven process” without having to invest the time to build out an extensive framework ourselves.” Patti Chan, VP Product Although Imperfect Foods’ development team is small, their impact is huge. They utilized Split’s integration with Segment to power experimentation across user flows on their site. Shortly after Imperfect expanded from produce to full grocery, the team tested a new feature to introduce grocery add-on subscriptions during signup. In addition to the starter produce box, customers could opt into receiving meats, grains, dairy, or even snacks. This feature made it even more convenient for customers to take advantage of the expanded product assortment, helping weekly order volumes to double since January. Tying back to its mission, Imperfect Foods is on track to save more than 200 million pounds of food from going to waste in 2020. The safety to take bold bets In another experiment, Patti notes her team “had a hypothesis that making some key changes to our subscription options during signup could radically increase the value we bring to our customers. With any new feature, there’s so many different ways to present the idea to your users. We tested a subtle version, a traditional version, and one risky version that we thought was very compelling but could add up to a minute in the signup process.” “Thanks to Split, we were able to learn that the risky version actually had the highest adoption by our customers while keeping overall signup conversion steady.” Moving faster in a time of crisis In early 2020, when COVID-19 drove most of North America into their homes and curtailed nearly all public activity, Imperfect Foods recognized a dramatic increase in the demand for their online grocery service that delivered fresh food to customers’ doorsteps. They were able to respond by increasing their volume to meet this demand, and still had interested customers growing their waitlist through spring. Shipping features with speed and efficiency Imperfect Foods also started using feature flags to make fast changes to their website. In one instance, the unprecedented spike in demand had led to a major delay in one region. Thanks to building an alert banner powered by dynamic configurations, the team was able to quickly show a message to just the affected customers without requiring any code pushes or deploys. Even non-technical product managers could make these feature changes, while engineers stayed focused on building tools to scale the business. This process not only increased developer velocity but drastically reduced the number of calls that their customer care team received. “Using Split enabled us to create the culture we wanted.” As Imperfect Foods continues to innovate in the grocery industry, Split is working with them to make sure they have the infrastructure and tooling they need to test and scale. Patti and her team are dedicated to experimentation-driven product development, and Split makes that possible. We’d love to have you try out feature flags with Split today, or talk to our sales team about our 14-day free trial to get started running your own experiments.'
Tags: experiments , experiment , startups , experimentation , data science , Imperfect Foods , imperfect produce , Data Analytics , product experimentation , product management , pm tools , a/b testing , lean startup , a/b tests , experiment design , how to execute on growth experiments , #productexperimentation , data-driven decisions , a/b/n tests , abtesting , feature flags
See also:
comments