Adarsh Mokashi

Product Designer

Adarsh Mokashi

Product Designer

Work

About

Resume

Contact

An AI-powered media recognition
product validated through lean experiments

An AI-powered media
recognition product validated
through lean experiments

experiments

An AI-powered

media recognition

product validated

through lean

Concept Validation

12 weeks

Feb 2025

Figma, Claude, Apps Scripts

Impact

Early validation results

Three lean experiments tested demand, sharing behavior, and follow-through. 67% of participants said they would use Sauce again, showing clear early interest despite a lightweight prototype.

Tests Conducted

Assumptions Validated

Validation Success Rate

67%

Context

People constantly discover clips without enough context to identify them

From TikTok edits to trailers and reposted scenes, people often know they want the source, but existing search tools break down with short-form, mixed-media input.

Sauce explores a faster path from curiosity to identification.
It is an AI-powered concept that recognizes movies, shows, and viral clips using audio and visual input, then helps users act on that information through cast details, streaming options, and recommendations.

The goal was not to fully build it, but to validate whether the behavior and demand
were real.

The goal was not to fully build it, but to validate whether the behavior and demand
were real.

Challenge

How might we validate whether this solves a
real user need?

How might we validate whether this solves a real user need?

The challenge was to validate whether users truly need an AI layer for media rediscovery or if existing search habits already suffice. Proving that gap was essential to justify building Sauce as a real product rather than a novelty.

Risky Assumptions

I identified and tested the top three assumptions that could make or break Sauce

I started with 14 assumptions and narrowed them to the three riskiest: whether people wanted to identify content they stumble upon, whether they would share clips from other apps, and whether they cared enough to continue into a “where to watch” flow. I tested these first before investing further in the product.

Scientifically, I confirmed that delaying caffeine allows adenosine levels to rise to a point where coffee can have a stronger effect.

This research shaped early design decisions, such as including progress tracking, recipe suggestions, and an onboarding moment that reframed the delay as a challenge instead of a punishment.

These assumptions guided all subsequent experiments and became the backbone of Sauce’s market and business validation.

Pretotyping Experiments

I designed three lean experiments to validate Sauce without building the product

Pinocchio Test - Testing post-identification intent

Pinocchio Test - Reinvented the Reel

Built a lightweight prototype using Figma Make with a “Where to Watch” CTA and tracked clicks through Google Scripts. Only 2 of 5 users clicked, suggesting that identification was more compelling than the follow-up streaming step.

Fake Front Door Test - Made a landing page, caught some people's interest!

Built a landing page with a waitlist to test interest before building the product. Out of 32 unique visitors, 6 signed up, resulting in an 18.75% conversion rate. This showed clear early demand for instant content identification.

Mechanical Turk Test - Testing willingness to share clips

Mechanical Turk Test - Got ghosted by a few participants :(

Recruited 5 participants and asked them to send TikTok, Instagram, or YouTube clips through WhatsApp. 2 of 5 shared clips, meeting the success threshold and validating that users were willing to send content into the system.

Insights

What I Learned from Testing Sauce

Three lean experiments revealed how people actually behave, not what I assumed. I discovered what sparked curiosity, what felt natural, and what did not, insights that helped me define Sauce’s real value before building anything.

Curiosity is the hook

The landing page showed that people are genuinely interested in a product like Sauce. Excitement exists before a product even exists.

Sharing behavior feels natural

Users were comfortable sending clips from TikTok, Instagram, and YouTube. This validated that Sauce could fit into existing habits without heavy onboarding.

Identification matters more than streaming

The prototype showed that users cared more about identifying content than clicking “Where to Watch,” proving that curiosity mattered more than convenience.

Validation clarified the product direction

By testing assumptions before building, I learned where to focus: prioritize instant recognition as the core value, and treat streaming links as a nice-to-have.

Next Steps

The findings helped me plan what to test next

If the project were to continue, these steps would help validate Sauce at a larger scale and uncover its business potential. Findings suggested that deeper validation should focus on long-term engagement and brand positioning.

Increase Reach

Promote the landing page through Reddit and social media to attract more visitors and verify if conversion rates stay consistent with a larger audience.

Scale Clip Sharing

Expand the user pool to 20–30 participants and automate message flow to see if people continue sharing clips over time.

Refine “Where to Watch”

Run A/B tests with new CTAs like “Stream now on Netflix” to measure if clearer wording drives
more clicks.

Test repeat behavior

Measure whether users return to identify more clips over time, not just whether they engage once.

Reflection

Curiosity alone isn’t enough.

Curiosity sparked the idea, but validation determined whether it deserved to become a product. By turning assumptions into small, measurable tests, I learned how to assess demand quickly, identify what users actually valued, and focus the concept around real behavior instead of speculation.