Deal3
Lock in Web3 deals between brands & communities.
Demo
About
The problem it solves **Deal3 tackles five pain-points that drain Web3 marketing budgets today** Agencies open negotiations at **US $5 k–$20 k** a month and still chase surface-level follower spikes that rarely translate into wallet actions. Roughly **US $1.3 billion** disappears to fake-follower fraud each year, while nearly half of buyers are abandoning tools that keep adding new subscription fees. On top of that, communities must trust brands to pay after the fact, even though results are hard to verify, so genuine contributors often see only a fraction of the value they create. - **Deal3 replaces that broken loop with a Base-native, pay-for-performance marketplace.** - **A flat 2 % success fee** will be deducted automatically from an on-chain smart-vault which eliminates monthly retainers and hidden overhead. - Our **interest-matching engine** shows each brief only to vetted DeFi, GameFi, or SocialFi communities, ending spray-and-pray outreach - Funds are locked in escrow; they **auto-release only when on-chain KPIs fire**, removing payment risk for communities - **Wallet-level analytics** confirm every participant is a real address (Base gas ≤ US $0.01), wiping out bot fraud at negligible cost - An **automated reward splitter** sends tokens or NFTs directly to contributor wallets, ensuring fair, transparent distribution with no agency skim By collapsing retainers into a single 2 % fee, verifying engagement on-chain, and guaranteeing instant, milestone-based payouts, Deal3 turns wasted spend into measurable results. Aligning brands, communities, and contributors in one clean, trust-minimised flow. Demo: [https://youtu.be/V2BtwNp_bEQ](Link) Pitchdeck Video: [https://youtu.be/3kVHShxVIew](Link)
How it's made
Challenges we ran into **Prompting in Ohara** - It is not working or responding properly. The fix I implemented was changing the Model Prompt to GPT 4.1. **Expanding Idea** - I attended Base Batch's workshop and learned insights and suggestions about the project I'm working on. **Regarding the GitHub repository** - I had trouble porting it to GitHub and making it work via localhost. I asked Ohara what files were needed to run locally and used Cursor.ai to fix the code to run smoothly.