my role
See what my team had to say about working with me! I worked alongside a lead product designer, content designer, lead product manager, and a team of engineers. I was responsible for leading and iterating on the designs.
context
Klaviyo, built for eCommerce marketing, has served brands like Fashion Nova and Skims. However, it lacks full support for non-eCommerce objects with schemas, such as pets, subscriptions, children, schools, etc. You can think of a schema as the column headers in an excel spreadsheet: it tells you what each field is named, and what kind of data it will hold, such as dates, string (text), numeric, boolean (true/false), etc.
problem
Businesses like vet offices need to track detailed information beyond basic customer profiles. They need to manage multiple pets per owner, each with unique characteristics. However, Klaviyo currently lacks the tools to define and organize this data, making it impossible to personalize communications effectively. This gap in functionality forces these businesses to seek alternative solutions.
Non-eCommerce customers are currently forced to use a clunky workaround: creating numerous custom properties for each pet, per profile(pet owner). This creates a massive, frustrating data management burden.
solution
Value for Klaviyo customers: Custom objects are flexible data packets that users can customize and use throughout Klaviyo. Klaviyo's custom objects simplify managing complex customer data, like multiple gift cards, or pet subscriptions per customer. No more workarounds. Businesses beyond e-commerce, like hotels and vets, gain clear data visibility and unlock personalized marketing, driving revenue growth.
Business value for Klaviyo: By attracting new customers in diverse sectors like hospitality, veterinary care, and travel, Klaviyo can drive significant revenue growth and solidify its position as a versatile, powerful marketing platform.
project kickoff
We kicked off the project with the team, covering user needs, project status, open questions, and metrics. We emphasized the ongoing learning process.
I summarized the requirements to be an (1) object manager that would define an object and (2) the ability to use this object in other parts of the product [campaigns, segments, profiles, flows]. Outside of these capabilities, we were unsure of the smaller details of each feature. This was an opportunity for design to lead.
timeline
To accommodate engineering's time to build out the infrastructure, the project was divided into phases. Design aimed to stay ahead of engineering, but we intentionally avoided full Q1 design completion of the object manager to allow for requirement changes. We planned further design iterations to the object manager in Q4, leading to the Q1 2025 launch of objects.
challenge
Building this new feature from the ground up presented unique challenges, including uncertainties in scope, user needs, and engineering feasibility. While a business priority, the project faced a long road ahead due to the lack of necessary infrastructure. We anticipated frequent scope changes, rapid iterations, and extensive cross-functional collaboration to ensure seamless integration across all product touch points.
Resolution: To proactively address these challenges, I initiated weekly design meetings with the team to facilitate in-depth design reviews and gather crucial engineering feedback on feasibility. Additionally, I proposed open-ended discoverability testing to gain a deeper understanding of user interests and gather valuable feedback.
user flow
To start designing, I created a basic user flow with the key steps that a user would need to complete in order to create an object.
discoverability testing
To assess the direction and potential of custom objects, I collaborated with a UX researcher to develop a prototype and research plan for discoverability testing. We recruited a diverse group of five customers, including those with varying levels of experience with custom objects, as well as a mix of marketers and engineers.
The prototype was intentionally open-ended to encourage exploration and gain deeper insights into user expectations and interests. This approach allowed us to:
Identify the most intuitive name for the feature: We wanted to understand how users conceptualized the functionality and what terminology resonated with them.
Evaluate the usability of the object creation workflow: We assessed the intuitiveness of the process and identified any areas for improvement.
Gauge user confidence: We measured how comfortable users felt using the feature and whether they had the information they needed to proceed.
Gather feedback on design elements: We sought input on what aspects of the design were effective and what could be added or removed.
The findings from this research were instrumental in shaping the initial release of custom objects, ensuring a comprehensive and user-friendly experience
research results:
Some users did not understand the necessity of the “choose template” step and most jumped right into the most intuitive step which is “mapping fields.
Low understandability of the term “Klaviyo object” in the “Create[feature name]” step.
Users did not find it important to view all object data in the objects page.
challenge
While discussing data ingestion with a designer from another team, I discovered we were designing for nearly identical problem spaces. Recognizing the potential for duplicated effort, I proactively organized a meeting with both teams and the respective product managers. This facilitated crucial scope alignment, preventing redundant work and streamlining our efforts moving forward.
usability testing / quarter break between previous iteration
After taking a quarter break from working on object manager, to address other priorities, we wanted to address the results from our discoverability testing and conduct another usability testing session. We tested with 3 users familiar with custom objects and 2 users who were not familiar with custom objects.
As our vision of the feature became more clear, we wanted to take advantage of these sessions and test both V1 and future scopes.
research results:
🎉All 5 users confirmed the flow was intuitive to use and that they understood custom objects.
hi fidelity prototype
As much as we wanted to build out everything in the usability prototype, with a near launch we could not accommodate all those changes. In the last months before launch, the prototype had to be changed a number of times due to scope creep and engineering feasibility.
handoff
I created an annotated handoff file with the different flows and edge cases such as errors, deleting an object, editing an active object, and editing a syncing object.
results
🎉As of April 2025, the feature has produced $1.2M in ARR.
Custom Objects is one of the most impactful features we can launch in 2025. We've got about 64 deals that are either at risk or blocked because we don't have Custom Objects yet, which amounts to about $1 million MRR in total.
