Decentralized Data Rewards
AI companies were harvesting user data for billions. Users got nothing. We changed that.
Every major AI model in the world was trained on data scraped from the internet — data that users generated, for free, without their knowledge or consent. Grass was the first platform to pay users for it. The design challenge: make people comfortable with something they'd been taught to fear.
Setting the Scene
AI companies need massive amounts of real-world web data to train models. The current system for collecting it is fragmented, expensive, and ethically questionable. Users have zero visibility into what's being taken from them and zero benefit from it. Grass built a decentralized data network that incentivized users to share their unused internet bandwidth — with radical transparency and a real financial reward.
The Problem
The product concept was solid. The design challenge was trust. Users had spent years being told their data was being harvested without consent. Now we were asking them to opt in intentionally and share their bandwidth. The obvious question: why would anyone trust this? The answer had to be baked into every screen of the experience.
What made it hard
- →Privacy concerns were the starting point — users were naturally suspicious about what bandwidth sharing actually enabled
- →Technical literacy varied dramatically: some users were deeply technical, others just wanted to earn money passively
- →Regulatory uncertainty — data collection and token rewards had unclear legal status in multiple jurisdictions
- →Blockchain complexity created real UX friction — crypto wallets and token mechanics were a barrier for most users
- →Competitive alternatives existed — needed to differentiate on transparency, not just reward size
Users weren't opposed to sharing data. They were opposed to being exploited. The distinction matters enormously. Radical transparency about what data was being collected, and fair compensation for it, turned a privacy concern into a value proposition.
What the Research Revealed
- →Privacy concerns were the #1 barrier — users wanted to know exactly which sites were included and exactly what data was being collected
- →Blockchain complexity was a significant friction point — users didn't want to learn about wallets, tokens, or gas fees to earn a few dollars
- →Real-time earning visibility was critical — users needed to see value accumulating or they assumed nothing was happening
- →Community and social proof mattered — seeing thousands of others in the network increased trust more than any explanation
- →Data usage clarity was essential — users wanted to know their data wasn't going to military applications, advertising targeting, or anything they'd find objectionable
The Approach
I designed a trust-first experience built on radical transparency. Clear onboarding explained exactly what data was collected and why. A real-time earning dashboard showed value accumulating live. Community stats (500K+ users, $X paid out, partner companies listed) provided social proof. Blockchain happened behind the scenes — users saw earnings in USD equivalent, never in token prices. And granular opt-out controls let users exclude any sites they were uncomfortable with.
Radical transparency about data collection
Users saw exactly which sites were visited, which requests were made, and how this data would be used. No hidden collection. The transparency that felt risky to show was actually the thing that earned trust.
Real-time earnings dashboard as the hero
Earnings accumulating in real-time — by session, week, and month. Watching the number grow was psychologically powerful. Daily active users increased 3x when we moved this to the primary screen.
Hide the blockchain completely
Crypto transactions happened in the background. Users saw USD equivalent, not token prices. Even explaining 'tokens that convert to USD' created friction. Pure dollar equivalent worked better for mass adoption.
Opt-out per site
Users could exclude any site from data collection. This felt like control — and it was. Giving users agency over exactly which data they shared made the whole proposition feel cooperative, not extractive.
Proving It Out
Beta tested with 50K early users on the transparency-first design. Regular trust surveys to track comfort with data sharing. Onboarding analysis to find exactly where users dropped off. Retention tracking to understand what predicted long-term usage vs. churn.
- →500K+ users onboarded in the first 90 days — dramatically above projections and a clear demand signal for data monetization
- →Real-time earnings dashboard increased daily session frequency 3x vs. the control group — watching earnings grow is genuinely motivating
- →Community stats prominently displayed increased conversion by 23% — social proof moved more users than any feature explanation
- →Users with 10+ sessions had 4x lower churn vs. 1-2 session users — the product earned long-term loyalty through consistent reliability
Beyond the Numbers
- →Proved that transparency could overcome privacy concerns — users trusted Grass precisely because we showed them everything
- →Demonstrated strong user demand for data monetization — people wanted to be compensated, they just needed a trustworthy way to do it
- →Generated significant media coverage as the 'platform that pays users for their data'
- →Created a precedent for decentralized data networks that prioritize user benefit over extraction
What I Took Away
Transparency beats security theater. Users cared more about honest explanation than sophisticated privacy tech. Real-time visualization made the earnings feel real and motivating — abstract 'you're earning tokens' language didn't. And hiding blockchain complexity wasn't dishonest — it was design. The underlying mechanism doesn't need to be visible for the value to be real.
- →Social proof (seeing others earning) builds trust more than any privacy policy — prominent user stats and total payouts mattered more than almost any copy we wrote
- →Granular opt-out controls increased user comfort significantly — agency over exactly which sites were included made the whole proposition feel cooperative
- →Even explaining 'tokens that convert to USD' created friction — pure USD equivalent worked better and wasn't dishonest
- →Hiding complexity isn't deception — it's design. Users don't need to understand blockchain to benefit from it
What's Coming Next
A data analytics dashboard showing users what their data was actually used for — closing the transparency loop fully. Tiered rewards based on data quality and uniqueness. Dispute resolution for users who disagree with specific collection. Geographic targeting based on regional data value differentials. Corporate partnership programs for B2B data sales.