Leading Prime video ads

How can a unified interactive ads framework enhance engagement for customers and value for advertisers?

My Role and team

I lead Prime Video’s Ad Tech and Monetization UX team, composed of UX designers, researchers, design program managers, and design technologists.

For this project, I drove the end-to-end UX strategy for Prime Video’s Interactive Ad Framework, establishing consistent interaction models that are intuitive for viewers and effective for advertisers. Through deep customer research and cross-org alignment, I delivered a framework that reduces friction, clarifies the ad value exchange, and serves as the foundation for all future interactive formats

Time period

11 months from project kickoff to v1 launch

Project Background

Prime Video

Prime Video is Amazon’s global streaming service, serving 315M monthly customers across 240 territories with thousands of movies, series, and live events available across devices. It delivers premium entertainment through personalized discovery, immersive playback, and seamless viewing experiences that connect audiences worldwide.

Prime Video Ads (PVA)

Launched in January 2024, Prime Video Ads is building a multi-billion-dollar business that enables continued investment in content while improving the ad-supported viewing experience. Its mission: make every ad interactive, and make them entertaining, inspiring, and useful.

THe problem

For prime video viewers

When Prime Video Ads first launched, disparate product teams were leveraging multiple interactive ad formats. Customers were seeing inconsistent ad presentations and interaction patterns across Prime Video creating a fragmented viewing experience that undermined seamless ad integration. Not only we’re we seeing low engagement rates across the board, the engagement was inconsistent across the different interaction patterns, signaling to us that this was not a great customer experience.

For Advertisers

Advertisers weren’t fully sold on interactivity yet - they are used to QR codes; remote click interactivity is very new. Most see it as a play for lower funnel metrics and not relevant for awareness or reach campaigns. They weren’t convinced our solution was worthwhile, and the ones that did try it saw the aforementioned low engagement, making it a tough sell.

For Amazon

Amazon and PVA believe in a future of all ads being interactive — it’s critical for us to be able to leverage Amazon’s unique differentiators in the streaming ads space. Without a scalable and holistic design framework, we wouldn’t be able to do it well and at the quality bar both viewers and advertisers expect.

The Cross-functional component

Addressing this problem required substantial alignment across multiple orgs - and most of the required cross-functional relationships weren’t established yet.

• The Core Player team established a player framework, with multiple future releases, prior to the ad product existing. Naturally their top priority was driving viewer needs.

• Amazon Ads org has historically worked on all interactive ad formats across any Amazon publisher and their top priority was driving advertiser needs.

• Prime Vide Ads (my team) - in the middle of the these teams, looking for that balance across sides

• The Shopping org was focused exclusively on endemic lower funnel conversion and selling sponsored merchandise.

Why it was important to do it when we did

1. Consumer trust and expectations. Amazon’s ability to deliver best in class online shopping experiences and are aware of high levels of advertising on other Amazon properties, so interactive advertising on TV is aligned with consumers’ expectations of Amazon

2. Competitive Advantage. Prime Video can uniquely leverage Amazon's unparalleled e-commerce ecosystem, creating a distinctive advantage in the streaming landscape that competitors simply cannot replicate.

3. Build viewer habits on TV. Customers have low expectations for TV ad experiences not because they can't be done well, but because they're unfamiliar with interactive ads from traditional broadcasting. We have an opportunity to establish and teach customers a new standard for TV ad interaction.

4. Allow us to move faster. Failing to build the interactive ads framework at this early stage of PVA's development will hinder our ability to make timely decisions and move faster as we scale our offering of interactive ad formats.

Our opportunity

Leverage Prime Video's components to create consistent, cohesive ad experiences that enhance engagement to meet customer expectations and enhance advertiser value. Utilize a clear, consistent call to action with room with to experiment

Top problems to address

1. Lack of cohesive interactive patterns

• IVA till date has had discoverability, predictability, and engagement challenges despite high compliance with ad moderation. This may hinder us in making ad interactivity second nature in the future as we look for more cohesion with the player framework.

• UI issues -  Edge-to-edge safe zones causes the IVA UI to get cut off sometimes

• Ad moderation - Due to the small size, edge-to-edge safezones, and multiple locations, ad moderation compliance rate for today's IVA is 85% which allows a majority of advertisers to use IVA. However, due to the issues outlined above, it is not driving much value for customer engagement.

2. Lack of player chrome integration

• Lack of cohesion with the larger PV viewer journey - Cosmetically it may look like PV design language to some extent but it doesn’t use PV interaction patterns/components + detached from player chrome

• None of the tech was talking to each other.

3. Lack of metadata

• The customer is not displayed any metadata to help them in their decision making. We knew from research this was critical if they were going to engage.

our Design process 

Design a consistent, scalable UI

We started wide and narrowed to the following three design variants.

Option A: Single-click center press

Option A addressed the three things we wanted to tackle:

1. Increased prominence with clear focus state and consistent placement with player interactions.

2. Space for metadata so the viewer can be confident in what they are engaging with.

3. A clear call to action (CA) - intuitive and focusable CTAs for single-click actions to reduce cognitive load for viewer about how to interact. Viewers are familiar with this call to action from retail.
   
   

Option B: Single-click press up

Option B extended the player team’s framework. The benefits were:

1. Consistency with player nudge component -these are used throughout prime video, primarily during live sports. Offers a one click CTA while preserving center click to play/pause content

2. Like option A, provides space for metadata

Option C: Two-click press up:

Option C was a variation of B, but required two clicks to add to cart. The benefits were:

1. Maintains consistency with the player nudge component, while allowing for scale when a customer may have multiple actions within an ad.

2. Provides space for meta data.

integrate into the player chrome

The original IVA ineractions existed as isolated overlays, disconnected from player controls and functionality. This created an either/or experience and made ad interactivity an outlier to all other playback interactions where they are built into the player chrome. We also knew there were future Play framework expansions rolling out across content types, starting with sports.

When we first started to explore how to integrate ads into the player chrome, the player team proposed we 1) Hide IVA CTA when IVA is not in focus and 2) Hide IVA component completely when focus is on player chrome. While these recommendations work well for player features like X-Ray, they did not work well for interactive ads especially since there is a limited time duration to interact with ads and ad interactivity on TV is still a relatively new concept.

Our decision: Show IVA at all times

We pushed to have the IVA CTA visible at all times and have the IVA component available at all times as the viewer navigates through the player chrome. To gain alignment, we conducted user research that ended up confirming our solution would work.

address the lack of metadata

IVAs relied on the creative driving the “what” behind the product or service for the viewer to take action on. Yet UXR with 15 customers revealed that all participants are frustrated about the lack of metadata, which makes it nearly impossible to know what exact product is being added to cart before adding it. We ran multiple user research studies and met with user research in the Shopping org to understand what meta data would be most helpful to a customer in a streaming setting.

validate with customers along the way

We conducted user research throughout the entire design process to help guide our decision making. Generally, we wanted to understand:

1. Usability and risks of each IVA option

2. Usability of interaction patterns holistically across PV

3. What meta data was important to encourage engagement with IVA

Key takeaways

When it came to the two different interactions, there was no perfect solution.

Center-Click IVA

• Preferred by ~⅓ of participants (7 of 21) for its simplicity and speed under time pressure.

• Risks:

  • Conflicts with the pause behavior—half of users who attempted to pause during the ad accidentally added to cart.

  • When customers acted unintentionally, they wanted clear undo or confirmation messaging to recover easily.

Press-Up then Center-Click

• Favored by the majority (14 of 21) for feeling deliberate and distinct from pause behavior.

• Risks:

  • Slightly slower in timed settings.

  • “Press up to shop” phrasing created expectations of a full shopping flow, which reduced engagement since viewers prefer quick returns to content.

Additional Insights

• 13 of 21 participants were satisfied with lightweight product info (stars, reviews, deal badges).

• 15/21 preferred IVA options visible throughout playback, aligning with our integrated player chrome solution.

• Customers prioritized consistency across ad interactions within a break—interaction method mattered less than predictability.

• While most appreciated the ability to act on ads, their main goal remained getting back to their content quickly.

secure Cross-org alignment for launch

With confidence in our designs, we we’re ready to work on our launch plan. At this point, we had relationships built across orgs, and teams were generally onboard with what we wanted to do. The next challenge was getting teams to align on how we would roll this out. There were two key decisions we had to drive:

Roll out with intention

Research showed us that viewers have low expectations of TV ad interactivity to begin with, and for ads specifically they expected press up to take them away from their playback experience so they hesitated to take the action. However, due to the conflict in pause behavior and based on viewer concerns about accidental ad actions, it was important to experiment with press up as well. Rather than making an immediate shift to new press up interaction pattern we proposed we

1. First establish a new baseline IVA that builds confidence through simplicity - maintaining the familiar single-click engagement that has historically existed while adding clear focus state

2. Simultaneously integrating this with the player chrome, creating a foundation for future enhancements while allowing viewers to learn they have a consistent place to access interactive content related to what is on screen in that moment.

This gradual evolution helps viewers adapt to interactive ads in a familiar context while maintaining their comfort with the viewing experience.

Shift how ad success is measured

All of the design, research, and cross-org collaboration on interactive ads led us to shift how we think about ad success measurement for Prime Video. Instead of solely upper funnel measurement like CTR and impressions, we added lower funnel metrics that are more valuable signals that measure intentionality: branded searches, detail page views, and attributed purchases - all within a 14 day period.

Launch designs

We worked with engineering to create a set of flexible, accessible templates that supported our new UI, interactions, and meta data options. 

Post-Launch results and adoption

The framework has rolled out across all four orgs, with 8+ teams using this in production to date. Collectively, we project our efforts will improve viewer’s ad experience (engagement has increased while also contributing to >$[redacted] in supply value by EOY’26. Since launch, we have seen +1,000 bps in engagement.