As design lead, I owned the redesign of Amazon's installment financing experience from strategy through launch — turning an initial 7-country rollout into one framework that now scales across 33 countries.
Short on time? Pick your read.
Full case study — process, tradeoffs, and craft.
On Amazon's Detail Page, installment offers lived in the old Twister+ swatches — parked far from the price and easy to miss. The framework capped coverage at 72% of eligible views and locked out whole product categories.
March 2022 research was blunt: customers wanted these plans next to the price, and they couldn't compare options to make a confident call.
Rather than patching Twister+, I led a complete rebuild: Unified Financial Offers (UFO) Priceblock. We surfaced the cheapest plan directly under the ASIN price and added a Compare Plans table to let customers weigh options in a single glance. By building a new framework independent of AUI, we opened the door for new banking partners and scaled coverage to 99% of eligible views.
As the day-to-day design lead, my role wasn't just owning the interaction model—it was driving the alignment to make it happen. I brought early solutions to our product, engineering, legal, and regional partners, navigating their competing constraints to find a common path forward. It was a constant balancing act between customer data, technical constraints, and tight timelines to earn buy-in across the organization.
None of it came for free. When an AJAX-rendered Compare Plans table—built to cut page latency—broke a one-click opt-in experiment, we hit a wall. I evaluated three engineering workarounds and ultimately recommended the path that cost us four extra weeks. It was a tough trade-off, but it was the only option that didn't compromise one group’s experience just to speed up another's.
Foundational and overseas discovery: I partnered with external research teams, including AnswerLab, and with our internal research partners on moderated qualitative studies. I defined the learning goals with the moderators, walked them through the product decisions we needed to unlock, observed sessions, and translated findings into design requirements.
Device-pattern validation: studies generally split participants between mobile and desktop/web so we could distinguish device-specific constraints from problems in the interaction model itself.
Agile course correction: between larger studies, I led smaller unmoderated tests for copy and interaction patterns and used lightweight usability feedback to adjust features before committing more engineering time.
Experimentation: weblabs tested information hierarchy in the comparison table and the prominence of Priceblock, including font-size treatments. The launched treatments balanced customer comprehension with sales, operating-profit, and conversion guardrails.
At the outset, I was the sole designer driving day-to-day development — a peer and I brainstormed early concepts together, but they left the team shortly after, and design partners overseas didn't join until our global expansion phase (more on that below).
By the project's full scale, the extended team included a Sr. TPM, 12 engineers, a QA engineer, 2 design support partners, 5 PMs, a UX writer, a visual designer, and a content strategist — plus 7 design partners across global regions and 2 banking partners once expansion began.
Some of that partnership was pure engineering trade-off. When the team moved Compare Plans to an AJAX-rendered table to cut Detail Page latency, it broke a one-click checkbox experiment that let customers opt into an offer without ever opening the table — the checkbox's logic depended on the table already being present in the page. I worked through three technical options with engineering (decouple the checkbox from the table, grey it out while the table loads, or disable AJAX for anyone in the checkbox experiment) and recommended decoupling, even though it cost 4 additional weeks, since it was the only option that didn't degrade the experience for one group of customers to speed up another.
Separately, a UFO feature that carries financing offers forward as a customer keeps shopping triggered a latency regression beyond our team's threshold — about 3ms against a 10ms allowance. I audited every image asset behind it — card art, provider logos, iconography — found card art and provider logos were the biggest offenders, and drove the work to standardize and shrink them, cutting the average payload by roughly 95% and clawing back the latency without touching the customer-facing design.
of customers viewing the table make a selection. This low initial conversion points to a major drop-off right at the start of the funnel.
of customers who click a plan proceed to final confirmation. This proves that the core friction isn’t the offer itself—it's getting them to take that first step.
Increase visibility of key benefits (lower APRs, monthly payments) above the fold to drive higher selection rates.
Tap a piece of UFO Priceblock, then switch lenses. The words change more than the facts do — that contrast is the point.
I want to know I can pay this off over time before I even think about adding it to my cart — not dig through swatches to find that out.
UFO Priceblock — the cheapest available plan surfaced directly under the ASIN price, built AUI-independent so it isn’t bound by Twister+’s constraints on banking partners or layout.
Eligible-view coverage went from 72% to 99% — the framework swap is what unlocked the jump, not a copy or visual change.
If there are multiple ways to pay, show me all of them side by side — don’t make me guess which is the good deal.
One Compare Plans table component, rendered per-market with whichever banking partners are eligible there — one comparison pattern instead of a bespoke table per country.
Offer selection rate rose from 14% to 20% — customers who could compare confidently chose a plan more often.
I don’t care what market I’m shopping from — I just want the same clear options everyone else gets.
Shipped mobile-first in the US, then set the pattern for a 7-country, multi-partner rollout — every component built to reflow (not truncate) for markets like German-length text from day one.
$331M in live annual impact today, with $900M in projected GMS across 33 countries as the rollout continues.
Shipped and live today on Amazon. See it on Amazon →
Comparison Table 2.0 is now live in 33 countries across North America, Europe, Latin America, the Middle East, and Asia-Pacific — Israel and Turkey among the most recent to onboard — with a projected $900M in GMS. To handle the expanding volume of offers in emerging markets, I’m running usability studies on new plan-layout variations. I’m also collaborating with our international partners to flag scalability issues early, catching architectural gaps before North American implementation even starts.
The real win wasn't a screen — it was building a system the business could scale on. Shipping this meant managing intense deadline pressure while refusing to let the user experience degrade for the sake of an engineering shortcut.
If I were starting this over, I'd bring Legal and Banking into the room during the framework design phase, not after — the France/Italy disclosure rules and Mexico's view-only requirement should have shaped the component API from day one instead of being retrofitted as market-specific exceptions.
Priceblock and Compare Plans run on a reusable component & motion library I built to scale across regions. Explore the UFO Design System →
Reaching 33 countries wasn't just translation and currency formatting — every specialist partner brought constraints that reshaped the design itself.
German character counts ran far longer than English inside the same fixed-width components. Fixed by designing every component to reflow, not truncate — one system, no per-market forks.
Both markets prohibit displaying complete banking term details on the product page. Fixed with a compliant summary view, full terms disclosed through the required legal channel instead.
Banking partners required the offer be visible on the product page but not selectable until checkout. Fixed with a "starting at" preview state, with plan selection moved to checkout.
Both markets use installments for roughly 75% of purchases, each with far more competing providers than a mature market. Fixed by scaling the provider picker itself without it becoming the bottleneck.
These trade-offs ran through every specialist partner I worked with, not just the four above. See who was in the room →
Scaling to 33 countries surfaced the same market-specific questions over and over — banking rules, disclosure law, provider density — so I wrote a wiki page documenting every trade-off and the reasoning behind it, and shared it worldwide so other teams facing the same expansion decisions didn't have to re-derive them from scratch.
That wasn't enough on its own. Docs answer questions people already know to ask; they don't surface the ones a market's own designers and stakeholders haven't voiced yet. So I raised it directly with my manager: I wanted to influence global expansion beyond my own team's roadmap, not just document it after the fact. I made the case for it, and secured business approval to travel to India, Brazil, and Dubai — sitting with local designers and stakeholders in each market to understand constraints no wiki page or status update would have surfaced.
A framework only scales as far as the people who have to live inside it understand it. Documentation gets you distance; showing up in the room gets you the truth.
We're turning "big purchases shouldn't feel big" into reality. By redesigning the Priceblock and comparison layouts, we surface transparent installment options right alongside the price—converting user hesitation into confident selections.
Charlie exhibited ownership by creating processes that elevate team quality and developing a comprehensive 3-year installments vision with personas and short/long-term roadmaps centered on "Choice in how to pay is as important as what to buy." He consistently delivered results through his feasible 2026 vision work and cross-team collaboration between the Installments and Shopping teams.
