Netflix House: Checkout 2.0

Improving ticket discovery, choosing and purchase for users, while testing into a new business strategy.


Role

Product Design Lead, House XD
Led a small team of flex designers

Category
Product Design
Design Systems

ScopeConsumer UI / UX, Tooling UI / UX, 3rd Party Integrations

Timing
Q1–Q2 2026

Key Stakeholders
Consumer Products & Experiences: Marketing, Operations, Revenue, CRM
Product
Engineering
After an MVP that launched offering one ticket per experience, Netflix House needed to test a wide ticket assortment, but our third-party ticketing platform limited the purchase experience. I proposed a hybrid approach: owned UI for discovery, cart, and checkout, with targeted third-party modules for specific selections. Conversion improved significantly, and new offers could launch without an engineering release.

Netflix House is Netflix’s physical venue for immersive attractions, themed dining and retail. The MVP ticket assortment was very simple, offering a single ticket selection per experience. The updated business strategy was to test into a wide ranging ticket assortment, to see what performed best across locations, experience types, and audiences. Checkout 2.0 was the next step in evolving the ticketing and purchase experience, with three goals:

  1. Create a more streamlined purchase experience for our users
  2. Integrate checkout with the branding and visual language of Netflix House
  3. Increase our ability to iterate quickly and learn through A/B testing

Because MVP tickets went on sale months before opening day, we had limited data on how far ahead guests would book or how they’d buy once they could book closer to their visit. Our first post-launch priority was to stay agile, experiment quickly with different ticket offerings, and learn our way toward the ideal state.




Design approach

We kicked off with stakeholders across marketing, product, operations and CRM to understand the wish list and must-haves. Early flow explorations helped facilitate discussions about potential paths forward for ticketing. Three principles guided my work:

  1. Lead with solutions over constraints. Operating in ambiguous, high-pressure environments requires acknowledging constraints while giving engineering and business partners viable "80% paths forward."

  2. Map systems early to move fast. Untangling dependencies and complexity up front prevents major pivots during implementation and allows informed decisions throughout.

  3. Great experiences start behind the scenes. Great consumer-facing experiences are impossible if internal systems slow things down, so the full end-to-end, including tools and workflows, matters.


Key decisions

  1. Go hybrid instead of owning the whole flow. The third-party vendor created real UI and UX friction for guests, but it also let us deploy and test multiple ticket structures quickly. Committing to a fully custom experience too soon would have slowed us down and tied up design and engineering time. I proposed a middle ground: owned UI for merchandising, the cart and checkout, with the third-party modal kept for quantity selection, date and time, and add-ons, deferring a fully owned build until our ticketing models matured.

  2. Keep users in the right context. We needed to own discovery and the multi-experience cart so users could plan their visit and merchandise tickets and promotions across pages. The third-party ticket comparison forced users deeper into the funnel before they were ready to make a decision, and required visual assets per ticket and block copy instead of a scannable bulleted approach. Inline ticketing in owned UI kept the narrative context and allowed scannable, side-by-side comparison of choices.

  3. Separate storytelling from selling. We launched an All Tickets page separate from the Location and Experience pages. It gave high-intent users a place to land and convert, and let the Location Landing page focus on explaining the Netflix House concept.

  4. Improve internal systems to improve the guest experience. Self-service ticket management meant offers could be managed and launched quickly without an engineering release. The influx of new ticket types demanded a structured offering hierarchy, so we built a dual-sided system: merchandising tools for marketing partners and high-clarity discovery surfaces for guests.

  5. Ship static cart recommendations first. We launched an MVP set of recommendation cards ("Don't miss out on these") that met the immediate discovery need, which let engineering scope the dynamic conflict-detection engine for a future iteration.




What happened

Conversion and speed to market: Conversion improved significantly, and the technical support needed for new ticket concepts stayed minimal, so new concepts were easier to launch than if we had taken on the entire flow end to end.

Multi-experience behavior: Netflix House moved from an isolated, single-ticket booking model to a full-itinerary experience, with measurable increases in cart size and completion.

Order recovery: Abandoned cart recovery links route guests back to their complete multi-item carts, so they can finish their full booking in one transaction instead of reassembling lost plans.

Add-ons and upsells: The hybrid approach made it possible to test add-ons and upsells in the 3P flow, which supported the decision not to build an idealized solution before we had more data.

Operational autonomy: Realtor CMS self-service campaign management decoupled promotional merchandising from engineering release cycles. Marketing and revenue teams could launch, test, and adjust offers in real time, and Product and Engineering could focus on platform architecture rather than routine operational support.



What I learned and what’s next

After launch we A/B tested the placement of the inline ticket modules, and the results validated my instinct that users need key details before they are ready to purchase from an experience page. Next steps included evolving the the foundation for multi-booking support in transactional messages, including full-itinerary order recovery messaging. If I did it again I’d:
Build for a much bigger catalog from day one.

We designed for the number of tickets we had at launch, and it grew fast. I’d set up stronger organization early, so the system could absorb growth without clutter.
Prioritize filtering and sorting sooner.

As the catalog grew, users needed better ways to narrow their options by date, experience type, price, or group size. I’d move those tools out of the "fast follow" bucket and into the launch plan.
Give promoted tickets more emphasis and more places to appear.

Featured offers appeared only on the All Tickets page; I’d add placements on other pages, like experience pages. Guests could see what’s featured in more places, and the business could spotlight more of what it wanted to sell.

Currently
Open to senior and staff product design roles, especially design systems and tooling. 
About this work
Some details and figures have been omitted or generalized for confidentiality. I’m happy to chat more in conversation!