Improving conversion with a streamlined checkout experience at PetSmart

Client

PetSmart

Role

Principal UX Designer

Timeline

Aug - Nov 2024

Platform

Web, App

Summary

PetSmart's web checkout featured a dated stepper design, and the app utilized an iFrame version of the web experience, contributing to increased abandonment rates. As the sole designer, I led a comprehensive redesign focused on streamlining the experience for users to expedite the time it took for existing users to checkout.

To accomplish this, I took this project through all stages of the UX process and championed conversations with stakeholders and technology to find viable solutions to reach the market sooner. Through rigorous user testing and iterations, the initial solution was implemented with future feature enhancements in product’s roadmap to further improve the experience.

Impact

Both the initial phased rollout, and iterations produced great results:

13%
Increased total conversion

11%
Increased sign ups

9.6%
Increased revenue per visitor

3 minutes, 44 seconds
Reduced time on task

$89 million
Increase in annualized revenue

77+
Net Promoter Score (NPS)

New web and app designs for mixed checkout

Problem statement

PetSmart’s checkout experience was underperforming market average, leading to increased cart abandonment rates. Users attempting to complete purchases within PetSmart’s web and mobile application were experiencing a frustrating and time-consuming checkout process due to an outdated interface lacking intuitive navigation and clear information. The interface did not adhere to Human Interface Guidelines or Material best practices, resulting in a clunky and unfamiliar user experience that hindered usability and eroded trust with pet parents.

Legacy web & app designs

Goals

Increase conversion rate

Reduce the amount of time spent in checkout

Reduce cart abandonment rate

Increase customer satisfaction and loyalty through brand perception of the digital experience.

Discovery & research

Very cumbersome, kept failing yesterday, too many buttons and other stuff to tap. The credit card expiry date doesn’t autofill from Google payments. Make it easier to spend my money with you!!
— Pet parent

Analysis of post-purchase customer feedback, derived from Medallia surveys and Contentsquare session recordings, revealed significant pain points within the checkout process for our pet parent users. Key areas of user friction included:

  • Inefficient flow for signed-in users: Existing signed-in users expressed frustration with the linear, step-by-step nature of the checkout process, indicating a desire for a more streamlined and personalized experience.

  • Excessive form fields: The number of required form fields during checkout contributed to user fatigue and potential abandonment.

  • Ambiguous error states: Non-descriptive error messages hindered a users' ability to effectively address and resolve issues, resulting in frustration and delays.

  • Difficulty applying loyalty points: Users encountered challenges in locating and utilizing loyalty points during the checkout process, suggesting a lack of intuitive integration and visibility.

These findings highlighted critical opportunities to enhance the user experience by optimizing the checkout flow, reducing cognitive load, improving error communication, and ensuring seamless integration of loyalty program benefits.

Competitive analysis of retail competitors checkout experiences

Competitive analysis

From the competitive analysis, it was clear PetSmart was underperforming in several areas of a modern checkout experience. Some of the main opportunities stemming from the analysis included:

  • An intuitive and streamlined flow to expedite user’s to place an order

  • Interstitial page for guest checkout and sign in

  • CTA placement prominence

  • API integrations with Google Payment

  • Geotargeting for autocomplete addresses & verified addresses

Requirements & recommendations

After identifying checkout issues, I collaborated with product and tech teams to prioritize improvements using the RICE model. We tackled immediate user frustrations with quick fixes while simultaneously redesigning the entire checkout experience for a better brand perception.

A key focus was simplifying the checkout for loyalty members, enabling a single-tap purchase to address the frustration with the existing multi-step process. This targeted approach aimed to maximize impact by focusing on the largest user segment.

Ideation & prototyping

Building upon insights from the discovery phase, I developed initial design concepts through the wireframing process to establish a streamlined checkout framework.

Prioritizing user efficiency and Medallia feedback, the design for both web and app shifted from a multi-step process to a consolidated review screen, enabling returning signed-in users to complete purchases with a single tap.

To ensure a consistent and intuitive experience, native app components were leveraged, minimizing technical debt. Furthermore, the primary call-to-action was strategically positioned at the bottom of the screen for optimal thumb accessibility.

App wireframes

Testing & iterations

Following the development of skeleton wireframes for layout, I created initial UI designs leveraging our design system, for user testing. The primary objectives of these tests were to validate:

  • Usability of the consolidated review screen

  • Assess users' ability to intuitively navigate, edit, & place an order

  • Reduce cognitive load within sections

  • Findability of topics within sections

Consolidating the stepper process

The interface was implemented with required content to meet product requirements. Steps such as the ‘About you’ and ‘Pick up in store’ were consolidated to reduce the amount of redundancy users encountered with duplicative fields. Additionally I added the dynamic disabling of the "Place order" button during edits to prevent errors.

Usability testing showed a 100% completion rate for guest users and a 90% success rate for signed-in users, with both groups finding the process easy. The app version, using native list components for a streamlined view, was also well-received, with users finding it intuitive and familiar to other retailers used.

Web guest view of Pickup method

App signed-in view of Delivery

Call to action placement

A comparative usability test was designed within UserTesting, employing an A/B testing approach, to determine the optimal call-to-action placement on desktop checkout. Option A presented users with dual call-to-actions, while Option B featured a single call-to-action positioned under the order summary.

The study, involving 15 participants, revealed that both design options were perceived as highly usable. However, 12 participants indicated a preference for Option B, citing its familiarity and alignment with their expectations based on established online checkout patterns, a finding consistent with Jakob's Law.

Call to action - Option A

Call to action - Option B

Option A button is the first thing you see, so it’s clearly easy. But every site that I shop at usually has the place order button under the order summary like Option B. So, my choice is Option B because thats where I’d expect to find the button.
— UserTesting participant

Reduce cognitive load in Veterinary section

The legacy veterinary section had a cumbersome form, requiring even signed-in users to re-enter pet data. The crucial authorization upload option was often missed, slowing approvals. Additionally, the continue button was confusing, leading to abandonment.

To fix this, field requirements were streamlined, and the information architecture was redesigned to clearly offer users a choice between uploading authorization or having PetSmart contact the vet.

Usability testing confirmed the improvements, with a 100% completion rate and 90% of participants finding the process very easy.

Legacy screen of Veterinary info

New screen of Veterinary info

Adding visual prominence to applying points

On the legacy website, users struggled to apply loyalty points. The initial redesign placed the redemption option in the left-hand section, but user testing revealed placement issues and a desire for incremental point application. Based on this feedback, the design was revised, moving the 'apply points' section above the order summary and adding incremental functionality. Subsequent testing showed a significant improvement in user satisfaction.

In the app, loyalty point redemption was entirely absent. To address this, a prominent redemption feature was designed and placed above the fold. Usability testing confirmed 100% successful application.

Following positive user feedback, the incremental redemption feature, along with its financial implications, was presented to Product and Loyalty stakeholders for review and alignment before moving forward.

Apply points within editable sections

Apply points within Order summary

Handoff & results

After positive user testing, the refined designs were prioritized for launch. To ensure smooth development, detailed specs and annotations were provided. Active participation in refinement discussions with Product and Development teams further ensured alignment with project requirements during sprint planning. Using design system components sped up front-end development.

A phased rollout (1%, then 50%) replaced A/B testing to manage risks. Significant performance gains (10-19% lift, 95% confidence) were observed, leading to a full rollout and an estimated $89 million increase in annualized revenue.

Addressing user feedback regarding pickup process friction, I redesigned key interactions. This included reducing cognitive load through field consolidation, enabling users to easily manage alternate pickup contacts, providing transparent and dynamic SLA information, and resolving ambiguity in multi-store selections.

Pickup checkout

Mobile Web Pickup screens

Desktop Web Pickup screens

App Pickup screens

I improved the same-day delivery user experience by focusing on clarity and efficiency. This involved redesigning time slot selection to provide greater transparency, reducing process steps by implementing a default to the soonest delivery time, and designing intuitive tip interaction patterns.

Delivery checkout

Mobile Web Delivery screens

Desktop Web Delivery screens

App Delivery screens

Reflection

The comprehensive redesign of this critical e-commerce touchpoint represented a significant undertaking, yet it offered a unique opportunity to profoundly enhance the pet parent experience.

This project established a strong foundation for future development, enabling the implementation of additional features informed by user feedback and research which did not make the initial implementation. Moving forward, I will partner with Product, who uses the RICE model, to strategically prioritize these feature enhancements and to ensure optimal user satisfaction and alignment with user expectations.

Previous
Previous

Building and scaling a design system at PetSmart

Next
Next

Reimagining a cruise booking platform for American Express cardholders