decrease in refunds
decrease in cancelled trips
Founded only 6 years ago in 2017, Lime is one of the fastest growing consumer apps in the world and has sparked a revolution within the ride-sharing market.
Limeâs rider team is a core driver of their top-line growth: they have the task of nurturing the companyâs expansion while fostering a sustainable and flourishing business model. Omar Guenena, an Engineering Manager who has been at Lime since 2019, emphasizes their vision: "We aim to establish ourselves as a reliable, cost-effective choice for individuals seeking daily travel solutions.â
This intricate task involves the perpetual creation and testing of various features across various user segments. As Omar elucidates, "Our goal extends beyond capturing the interest of transient users like tourists or joy-riders; we strive to craft an experience that converts them into engaged users, fostering sustained growth in our user base.â This team is a high-performing cross-functional unit of 40 stakeholders from Engineering, Product, Data Science and Global Operations. A crucial focus for them is ensuring the development of features that genuinely delight users while advancing business metrics.
A robust experimentation culture had always been part of Limeâs DNA, guiding their strategic product decisions with data. However, before adopting Statsig, self-service functionality was absent. This meant that each experiment had to be manually overseen by the data science team, introducing bottlenecks. This dilemma initiated their exploration of experimentation tools as they sought to overcome this hurdle.
Omar said that they needed to revamp their experimentation process as the company was seeing unprecedented growth and their existing resources wouldnât have sufficed. âWe would have faced scalability challenges if we had needed an equal number of Data Scientists as Engineers.â
Lime also contemplated the possibility of constructing an internal experimentation platform versus procuring one, yet promptly recognized that the latter was the more prudent choice. âWe donât have the bandwidth in house to develop an experimentation platform â which isnât our domain, and we couldnât risk affecting the business if something went wrong.â Limeâs ambition for exponential growth demanded a platform that could empower cross-functional teams to operate swiftly.
Lime evaluated over 6 different vendors and chose Statsig for a variety of reasons. Omar and his team were excited to partner with Statsig for fueling their long-term growth because of the pace at which Statsig was adding new capabilities to its platform.
âEvery time I spoke to Vineeth and the Enterprising Engineering team, they would show me new features they built. We realized that Statsig was investing in the right areas that will benefit us in the long-termâ, said Omar when asked what motivated him to pursue a partnership with Statsig.
âWithin a week of signing the contract, we ran our first experimentâthat was very cool!â
With Statsig, Lime was able to accelerate their experimentation velocity and began rolling out a range of innovative features that drove top-line growth.
In order to empower first-time users to have a seamless experience, Lime developed a âhow to take your first tripâ tutorial and ran an experiment to study its impact on refunds requested by first-time riders. This became a winning feature, driving down the refund rate by 20%, and significantly impacting top-line metrics.
Similarly, Lime developed and tested a simple, elegant gif that showed users how to kick the scooter stand to start the scooter and that empowered users to seamlessly start their ride and reduced short or canceled trips (trips that end prematurely) by 5%.
With its comprehensive full-stack experimentation capabilities, Statsig swiftly became an essential component of the companyâs initiatives, enabling diverse teams to comprehensively test and validate their features before rolling them out.
IoT teams jump on board: Statsigâs usage soon went beyond the rider team as the vehicle team started running experiments on devices (bikes). They now test new firmware updates by comparing a test group of bikes with older bikes to validate the impact on vehicle metrics such as TVD (trips per vehicle per day). Omar highlights the technical nuances of running experiments on the app vs vehicle, âWhen conducting an experiment within an app, obtaining a configuration from the cloud and modifying the app experience is straightforward. However, for bikes, the process varies, requiring distinct methods to assign bikes to specific treatments or configurations.â
Marketing/Growth teams get better visibility into the performance of campaigns: With Statsig, growth teams were empowered to experiment the effects of serving coupons to users. They were able to understand the impact of pricing changes on user metrics and the impact of coupons and passes on retention.
âFor each of these metrics, we ran experiments to ensure that the features we built were aligned with our goals.â Omar gives an example, âThe growth team looks to optimize pricing by changing prices and studying how users respond. In this case, âbookingsâ would be the primary metric, while ânumber of tripsâ and âretentionâ would be secondary metrics.â In this approach, Lime would start with an objective and work backward to design an experiment, maintaining a continuous evaluation of the impact of their features.
Lime used Statsig to test and rollout several winning features such as:
The path of experimentation opened up opportunities to innovate and tackle Limeâs key business issues from new angles. Omar shared instances of inventive features that Lime tested with Statsig, saying, âIn enhancing parking, we explored the use of AR to scan the surroundings and evaluate its impact on the parking experience. Does it expedite parking times? Are there any glitches? Likewise, with Lime Prime, does the promotion of a pass stimulate increased trip activity?â
Knowing answers to such questions allowed Lime to analyze the impact of their features and make informed product decisions. The team has also been actively engaging on Statsigâs slack community. âWeâve taken a lot of advice on experimentation from the Slack channel and that has helped us coach engineers on the team to run better experimentationâ, says Omar who is very proud of how much the team has grown in the last couple of years. âInitially only a handful of members had the experimentation mindset on my team, but it is great to see that the entire team and other teams are also doing it now.â
Lime has evolved into a mature phase of experimentation, conducting numerous tests to inform a range of product and business choices. Reflecting on Limeâs substantial experimentation expansion, Omar notes, "From waiting for a couple of months to get data for a test, we are now at a point where we test every little thing before rolling it out and measure impact in near real-time.â
For the next phase of growth, Omar wants to strengthen their experimentation culture by ensuring every experiment has a hypothesis. âIs there an expected change that will be big enough to justify running an experiment? Do we have the right metrics for this specific experiment? Weâre now focusing on the quality of the experiments we run, not just quantity.â
âIâm always a fan of the new features rolled out by Statsig and Iâm looking forward to all the new ones to come!â
Founded in 2017, Lime is the largest and most experienced micro mobility operator in the world. They serve more than 250 cities in over 35 countries, deploying more than 250,000 e-bikes, e-scooters, and adaptive vehicles every day. Lime riders have taken more than 400 million rides â 99.99% of which have been incident-free globally â saving over 3.1 million tons of CO2 from the atmosphere.