Supermetrics purchase journey
Supermetrics streamlines your marketing and sales data from all platforms into a single source of truth, providing accurate, real-time insights to help you make smarter decisions. This project focused on optimising the purchase flow to match the new subscription model.
Project overview and purpose
I worked on the redesign of Supermetrics' purchase flow to create a more user-friendly and flexible subscription experience. As Senior Product Designer, I spent 8 months on this cross-functional project from ideation to deployment which resulted in a 22% increase in ARR and improved conversion rates for paid self-serve customers.
Challenge:
The existing purchase flow faced three critical issues:
Complex product selection process that confused customers about their data source needs
Limited flexibility with single-destination restriction and difficult upgrade paths
Heavy reliance on sales team intervention, forcing customers into a permanent sales-led journey
Project Goals:
Enable flexible package customisation to prevent customers from paying for unused features
Implement multi-destination selection without sales team involvement
Streamline the upgrade/downgrade process
Implement transparent pricing structure
Research
Secondary research: current flow and customer feedback
Mapping of current destination purchase flows:
After analysing the existing purchase flows, we discovered that the disjointed system wouldn't effectively support our planned subscription model. We identified key areas for improvement and moved forward with designs and customer interviews.
We based our customer research on the hypothesis that users would prefer the new package model, thanks to its increased flexibility, customisation options, and better-aligned pricing. We also identified potential risks, such as failing to attract new customers if we didn't clearly communicate pricing or demonstrate sufficient value relative to cost.
For our customer research, we interviewed 10 customers across various segments, as shown below:
Through ten interviews, we identified four key themes that validated our assumptions and revealed opportunities to address key risks. I used affinity diagramming alongside Vertex AI to analyse the interviews and identify additional themes and insights.
Theme 1: Package Arrangement
Customers responded positively to the new package structure, appreciating its increased customisability and flexibility. Many customers highlighted their heavy reliance on TikTok as a connector. They were particularly pleased with the option to access TikTok in either of the first two packages, as it had previously been inaccessible to some customers.
Theme 2: Package Pricing
All customers were happy to purchase a package based on what they saw. 2 out of 10 customers noted that transitioning from the lowest tier to the middle tier package would require them to build a business case for company approval. Agencies, in particular, recognised the value of the middle tier package. They expressed that it’s increased number of accounts and users was a significant factor in their decision-making process.
Theme 3: Package Information Clarity
All six customers expressed confusion about the data source modal, citing issues with the category patterns and the lack of a consistently visible search bar.
Key opportunity
We observed that 6 out of 10 customers were aware of helpful features that weren't available in the self-serve package or trial version. Making these features available would strengthen our package positioning.
Design and development
During the initial design phase, I collaborated with our Principal Product Designer and another Senior Product Designer to develop the new purchase concept. After that, I took ownership of the project, leading the research and working with the team to implement the new flow.
We moved the purchase flow into the hub and simplified the package customisation process. We created user flows, mockups and conducted usability testing using Maze.
Based on usability testing through Maze, we found that while the flow was successful overall, it needed refinement in specific areas such as data source selection.
