Checkout

Project Overview

Throughout my tenure at Flow Commerce (acquired by Global-e in January 2022), a significant portion of my time dedicated to researching and designing a new cutting-edge international checkout solution. 

The main goal was to replace the existing checkout (utilised by over sixty merchants) that suffered from a declining conversion rate and increased page load time. This was primarily due to the absence of guardrails, allowing merchants to add excessive customisations using their own HTML and Javascript.

My Role

Joining the checkout team during its Alpha stage as the Senior Product Designer and Researcher, I: 

• Conducted comprehensive research to uncover issues with the checkout and identify areas for optimisation.
• Communicated with merchants already using the new checkout solution and those yet adopted it to gain insights and understand their requirements.
• Collaborated with the team to challenge findings and prioritise updates and new features based on maximum impact.
• Designed solutions to address identified issues and boost the list of checkout features.

Problem

Flow Commerce identified a pressing issue with their international checkout solution, as the page load speed had been steadily deteriorating over time, resulting in negative effects on conversion rates.

The root cause was identified as the additional HTML and JavaScript added by merchants to customise the checkout appearance and build their own features. These customisations not only led to slow page load times but also caused delays in development and updates, requiring extensive quality assurance efforts across all merchants.

Business Goals

Recognising the detrimental impact of slow page load times on conversion rates and the challenges posed by complex release processes, the business initiated a strategic effort to create a new checkout solution. 

The goal was to strike a balance between speed and customisation, ensuring a fast and seamless checkout experience for customers. At the same time, enable merchants to customise their checkout to align with their branding and user preferences, eliminating the need for custom code. 

This approach aimed to improve conversion rates and expedite the implementation of enhancements and bug fixes across the entire merchant base.

User Needs

Customers require a checkout process that is simple, clear, and secure.

Merchants need a checkout process that is highly converting, easy to manage, quick, intuitive, and user-friendly. They also expect the checkout design to align with their brand identity.

My Approach

Research

• Utilised session replay tool Fullstory to observe user behaviour and identify checkout flow issues and frustrations.
• Communicated with merchants to understand the custom features and branding elements they implemented that positively and negatively impacted conversion rates and user experience.
• Analysed data to identify user engagement patterns and areas where users dropped off during the checkout process.
• Gathered insights and recommendation from reputable sources like the Baymard Institute.
• Conducted unmoderated user tests using Usertesting and UsabilityHub.
• Maintained a backlog of research ideas and insights in ProductBoard for further exploration, design, and testing.

Design

• Leveraged UsabilityHub to gather quick feedback on alternative designs and approaches.

• Heavily utilised UsabilityHub's country-specific user sourcing feature to target feedback from users in specific countries, considering the international nature of the checkout.

• Prioritised features and opportunities using a RICE scoring system (reach, impact, confidence, effort) in ProductBoard to track insights and facilitate team prioritisation.
• Conducted A/B tests to measure the impact of design changes and iterate on the checkout solution based on data-informed insights.

Challenges

Limited direct access to customers: Due to the nature of the setup where customers technically belonged to the merchants, direct communication with customers was not possible. To overcome this, I heavily relied on watching Fullstory session replays and analysing data to identify pain points and areas for optimisation.

Initial lack of an A/B testing process: When I joined the team, there was no established A/B testing process. Recognising its importance, I actively advocated for the implementation of A/B testing to gather data-driven insights and optimise the checkout solution.

Updates could have a negative impact on some merchants: Updates that resulted in an overall positive uplift in conversion rates for the collective set of merchants could have negative effects on certain individual merchants. To address this, we would identify potential reasons for the negative impact and assess the impact on the merchant's conversion rate over multiple updates. The focus was on ensuring that there was still an overall positive trend in the merchant's conversion rate despite temporary fluctuations.

The existing checkout has been in use for an extended duration, leading to a reduced number of bugs: During my analysis of Fullstory session replays, I not only identified opportunities for potential improvements but also uncovered bugs that were impeding customers from completing the checkout process.