# Picking the three alternatives Mapping who exists is an input; positioning needs only the three alternatives the ICP actually weighs, and the one dimension you beat each on. ## Where to look, in order 1. **What the ICP does today without you.** The manual workaround, the spreadsheet, the multisig, the CEX, the Discord bot. This is usually alternative #1 and the one founders leave out. 2. **Solana projects doing the job.** Check live usage, not landing pages: the program's recent transactions, DefiLlama if listed, GitHub activity. 3. **The same job on other chains**, if the ICP can move there. For an integrator, an EVM version they would have to bridge to is a real alternative. 4. **Dead projects** in the space. Why they died is the bear case your positioning has to answer. 5. **Hackathon entries** ([colosseum-copilot](https://github.com/ColosseumOrg/colosseum-copilot/blob/0453ffe26e8d245152619fcc949689a6adaab1c6/skills/colosseum-copilot/SKILL.md) for crowdedness). They tell you how crowded the idea is, but they are rarely what a paying ICP compares you to. For each candidate, capture: name, link, live / beta / dead, chain, a usage number with its source, what it does well, where it is weak. ## Choosing three Choose the three the ICP would name if asked "what would you use instead?", and include at least one non-product alternative (doing it by hand, or not doing it). Then, for each, write: - **We win on:** one dimension, concrete enough to demo or measure: settlement time, a fee in bps, the number of signers, custody, the steps to integrate. - **It wins on:** be honest. The reader already knows, and leaving it out costs credibility. If you cannot name a dimension you win on for an alternative, that is a positioning problem, not a copy problem. Send it back to [idea-sprint](../../idea-sprint/SKILL.md). ## What makes the win stick Six moat types, with the question that tests each one in crypto: | Moat | Test question | |------|---------------| | Network effects (liquidity, two-sided markets) | Is there a credible path to the critical mass where it starts to compound? | | Switching costs | What would a user lose by moving to a fork? Composability makes this weaker onchain than in SaaS. | | Data | Does the product get better with usage in a way a fork cannot copy from chain data? | | Technical complexity | How long until a good team replicates it? The code is usually public. | | Distribution and ecosystem lock-in | How many wallets, protocols and tools depend on you, or route through you? | | Brand and trust | Would the ICP trust a no-name fork with funds? | Prefer a win backed by a moat. A win on price alone ends when a fork lowers its fee.