Shoppable Advertising Across Platforms
A multi-study applied research program examining how shoppable advertising is interpreted, tolerated, and acted upon across social media and streaming TV contexts, with direct implications for product strategy, interaction design, and monetization decisions.
At a Glance
- Led a four-phase, end-to-end research program spanning qualitative discovery, controlled experimentation, and behavioral validation
- Tested shoppable ads on social media, streaming platforms, and a simulated living room TV
- Collected and integrated multimodal data (behavioral telemetry, self-report, biometric signals)
- Developed the MEVN decision framework to support product and design teams in evaluating when shoppability adds value versus friction
What Problem This Program Addressed
Shoppable advertising is increasingly common across platforms, from social feeds to streaming television. These formats promise reduced friction between exposure and purchase, but they are costly to build and introduce technical and experiential complexity.
Despite their growing use, there is little consensus on whether people actually want to interact with shoppable ads, when they find them helpful, and when they experience them as intrusive or disruptive. Industry teams frequently face uncertainty about whether shoppability increases value or simply adds friction. This uncertainty creates real product risk: teams must invest in complex technical infrastructure without clear signals about user acceptance, conversion tradeoffs, or long-term platform health.
This research program was designed to answer a behavioral and emotional question at the center of this uncertainty: how do people make sense of shoppability, and how does that interpretation change across platform contexts such as social media and streaming TV.
Program Overview
The program was structured like a product research roadmap, moving from discovery to validation to synthesis.
Interview Study
Qualitative interviews to understand user mental models, expectations, and emotional responses to shoppable ads across platforms.
Online Experiments
Controlled experiments testing shoppable versus non-shoppable ads with behavioral telemetry, attention measures, and outcome metrics.
Living Room Lab
High-fidelity behavioral validation in a naturalistic viewing environment to test ecological validity.
MEVN Model
A conceptual and practical framework synthesizing findings into actionable guidance for design and product decisions.
Phase 1: Qualitative Sensemaking
The program began with exploratory interviews to understand how people encounter and interpret shoppable ads in everyday life. I led a small team that conducted interviews with participants across age groups to capture a cross-sectional view of technology adoption and platform use.
Interview transcripts were collaboratively cleaned and coded for emergent themes. Rather than starting with predefined assumptions, this phase focused on identifying how expectations, motivation, and perceived intrusiveness surfaced organically. This phase functioned as generative UX research, surfacing user mental models and expectation mismatches that would later inform experimental design and hypothesis testing.
The resulting themes were synthesized into an insight framework mapping when shoppability creates value versus friction:
| Theme | What It Captures | When It Helped | When It Hurt | Example Quotes |
|---|---|---|---|---|
| Novel | Surprise and curiosity driven by encountering a new advertising format | Early exposure, especially on TV where shoppability was unexpected | Novelty faded quickly once the format became familiar | “I didn’t even know you could buy things straight from the TV. I remember thinking, wait, that’s kind of cool.” |
| Informative | Perceived usefulness of product information or discovery | When ads surfaced relevant or niche products tied to existing interest | When information felt irrelevant or forced | “I like it when they show outfits from a show. Sometimes you see something and want to know where it’s from.” |
| Convenient | Reduced effort required to learn about or access a product | When users were already in a browsing or information-seeking mindset | When convenience crossed into feeling pushy or unavoidable | “It’s convenient because you don’t have to leave the app, but it can also feel a little pushy.” |
| Entertaining | Engagement and enjoyment from optional interactivity | On streaming TV where ads were already expected | When interactivity disrupted immersion rather than enhanced it | “If I have to watch an ad anyway, at least this gives me something to do.” |
| Intrusive | Perceived disruption of platform flow or user goals | Rarely helped; sometimes tolerated when aligned with platform norms | When shoppability clashed with relaxation or social use | “I’m on social media to see my friends, not to buy a new fridge.” |
| Deceptive | Confusion between ads and organic content | Rarely perceived as positive | When users accidentally clicked on disguised ads | “I didn’t even realize it was an ad until I clicked it.” |
Participants frequently described shoppable ads as surprising and novel, particularly on television. At the same time, many expressed discomfort when these ads interrupted moments of relaxation or social flow, especially on social media platforms.
Phase 2: Controlled Platform Experiments
Insights from the interviews informed the design of controlled online experiments comparing shoppable and non-shoppable ads across social media and streaming contexts. I collaborated with a faculty advisor and a technical partner to build a custom experimental platform capable of injecting ads at precise moments and recording interaction telemetry. This platform enabled precise manipulation, randomized assignment, and fine-grained interaction logging, supporting causal inference rather than purely observational analysis.
To ensure generalizability, we tested multiple product categories and formats rather than optimizing for a single vertical.
Shoppable ads increased attention and engagement in both contexts. However, increased engagement did not reliably translate into higher purchase intent. On social media, shoppability often reduced purchase intent, while on streaming platforms it increased it.
Phase 3: Living Room Behavioral Validation
To test whether these patterns generalized beyond online settings, I designed and implemented an in-person living room simulation using a mock TV interface and a physical remote control. Participants watched a television episode with embedded ads under randomized conditions.
The setup combined eye tracking, facial expression recording, interaction logging, post-experience surveys, and follow-up interviews. Each session lasted approximately one hour and required extensive coordination and custom tooling. This phase prioritized ecological validity over scale, trading sample size for realism to capture behaviors that are often missed in online-only studies.
Service Blueprint: Shoppable TV Interaction
This service blueprint highlights breakdowns between user intent, system prompts, and perceived interaction cost.
| Ad Appears | Attention Shift | Moment of Hesitation | Non-Interaction | After the Ad | |
|---|---|---|---|---|---|
| Touchpoints | Non-skippable ad break with shoppable overlay and prompt | Visual cues, animations, and “Press OK to learn more” | Remote in hand, overlay still visible | Overlay times out or is ignored | Return to program content |
| Customer Actions | Looks at the screen when the overlay appears | Glances between TV, remote, and phone | Considers clicking but pauses | Does not click; shifts attention to phone | Continues watching or scrolling on second screen |
| Customer Experience | Curious and mildly intrigued | Alert but non-committal | Uncertainty about what will happen next | Relief at avoiding an unclear action | Ad quickly fades from memory |
| Pain Points | Shoppability is unexpected in an entertainment context | Competes with entrenched second-screen habits | Unclear cost of clicking (pause, checkout, disruption) | No immediate payoff for interaction | Product interest does not translate into action |
| Opportunities | Frame prompts as optional and low-pressure | Acknowledge and integrate second-screen behavior | Make click outcomes explicit and reversible | Offer delayed or mobile-based follow-up | Shift from interruption to invitation |
Results replicated earlier findings. Shoppability increased attention and self-reported purchase intent, yet actual interaction remained rare. Interviews revealed hesitation driven by fear of irreversible commitment and reluctance to leave the viewing experience.
Phase 4: The MEVN Model
To synthesize findings across studies, I developed the MEVN model. The model explains shoppable ad performance as the interaction of four components: user motivation, platform expectation, perceived value, and network effects.
Shoppability succeeds when these components align. When they conflict, engagement may increase while conversion suffers. The MEVN model functions as a decision-support tool rather than a post-hoc explanation, helping teams anticipate failure modes before deploying shoppable features.
Product & UX Implications
- Align shoppability with why users are on the platform
- Reduce perceived commitment costs through reversibility, preview, and deferred interaction
- Make outcomes of interaction clear and reversible
- Avoid interruptive shoppability in relaxation-driven contexts
Why This Matters
This research program demonstrates that shoppability is not inherently good or bad. Its success depends on how well it fits user motivation, platform norms, and momentary expectations.
By translating an ambiguous industry question into a structured research program and a usable model, this work offers teams a practical framework for designing shoppable experiences that feel aligned rather than disruptive. For product, UX, and applied science teams, this work demonstrates how combining qualitative sensemaking, controlled experimentation, and behavioral validation can de-risk complex interactive ad formats before large-scale rollout.
