The crypto market produces new projects every week, with most eventually becoming worthless and a small number becoming genuine long-term winners. The problem is that early on, they all look roughly the same — polished websites, hopeful whitepapers, active Twitter accounts. Separating good projects from bad ones takes not luck but a disciplined research framework. This article gives you a due diligence process you can repeat on any project — from rapid screening to deep evaluation — in four steps.
The goal of Step 1 is quickly determining whether this project is 'worth spending more time on.' Most projects can be filtered out here.
Team background: Do founders and core developers have public, verifiable identities? Look for LinkedIn profiles, GitHub history, past employment records at known institutions, and past crypto project records (what they built, and how those projects are doing now). Anonymous teams aren't necessarily bad (Satoshi was anonymous), but they require stronger technical verifiability as an alternative trust foundation — open-source code and on-chain verifiable data. If a project is both anonymous and has no verifiable technical work, skip it.
Problem definition: What problem is the project trying to solve? Does this problem genuinely exist? Does solving it require a blockchain (or would a regular centralized database suffice)? Much of the crypto industry is solving problems that don't require blockchain, or attempting to build a fourth identical product in a market already dominated by established players.
Competitive landscape: Who are the existing players in this space? How is this project different? If the answer is 'same as X but faster and cheaper,' quickly ask: why would existing users migrate from a mature competitor? Network effects often make 'faster and cheaper' far insufficient in DeFi-type use cases requiring liquidity.
Funding background: Which VCs or institutions participated? What were the funding rounds and valuations? Participation from top-tier firms (Paradigm, a16z, Multicoin, Coinbase Ventures) is an important signal (though not a guarantee). Complete absence of institutional investment isn't necessarily bad, but institutional participation from unknown small VCs warrants further verification.
If Step 1 passes, the token structure is the second checkpoint. A technically excellent project with poor Tokenomics design is still a bad investment.
FDV / Market Cap ratio: Open CoinGecko or CoinMarketCap and check the gap between circulating market cap and FDV. Ratios above 5× require understanding 'at what point and at what pace future unlocks will enter the market.'
Distribution structure: What is the combined VC + team allocation percentage? How long are their lock-up periods? Check the project's official documentation (Tokenomics page, Whitepaper) or Token Unlocks (token.unlocks.app). How deep was the private sale discount? If the private round valuation was 10% of current market price, every early investor token has a 10× exit premium — post-unlock selling incentive is extremely strong.
Upcoming unlock events: Are there major unlocks (exceeding 10% of current Circulating Supply) in the next 1–3 months? If so, consider waiting until post-unlock to evaluate entry, or set your holding period to end before the unlock date.
Token's actual utility: What function does the token serve within the protocol? Pure governance (voting only)? Fee burning mechanism? Required for accessing protocol functions? The more functions and the more mandatory the usage, the stronger the token's value support.
Whitepapers and official documentation can say anything. On-chain data can't be faked. The goal of this step is using real on-chain metrics to verify whether the project has genuine usage.
TVL (if applicable): For DeFi protocols, check DeFiLlama for TVL absolute value and trend. Is TVL rising or falling? Is it organic growth (real users depositing) or 'mercenary capital' attracted by high APY liquidity mining (which will immediately withdraw when APY drops)?
Daily active addresses and volume: Check Daily Active Addresses/Users and daily transaction volume trends on DeFiLlama, Dune Analytics, or the project's own data page. A protocol with real users should have a relatively stable baseline usage even in bear markets — not complete collapse to zero when incentives end.
Large wallet behavior: Use Nansen or Arkham to examine major holder addresses (top 20) — are they increasing or reducing positions? Smart money movements (historically strong performers — institutions or large traders) are a valuable signal, though not a sole basis for decisions.
Code audits: Has the protocol's Smart Contract been audited by reputable security firms (Trail of Bits, OpenZeppelin, Certik, Halborn)? What high-risk or critical vulnerabilities did audits identify, and have they been fixed? A protocol without an audit is essentially asking you to put money into a black box that hasn't passed any security testing.
The first three steps gather information; Step 4 converts information into decisions. Use a simple scoring framework — rate each dimension 1–5: Team credibility (1–5), Problem authenticity and competitive position (1–5), Tokenomics structure (1–5), On-chain usage and security audits (1–5). Total score 16–20 = worth building an initial position; 12–15 = watchlist, wait for more data; 8–11 = high risk, consider only with minimal position; below 8 = skip.
Equally important is setting Exit Criteria before entry: Why are you buying this project (what is your thesis)? Under what conditions is your thesis invalidated (e.g., TVL declines more than 50% for three consecutive months; a major competitor announces a feature copy and launch; a core developer departs)? People who set exit conditions in advance can actually execute rational exits when those conditions are triggered — rather than letting emotions decide to 'wait and see.'
This due diligence framework can't guarantee that your chosen projects will succeed, but it effectively filters out most obvious traps. In crypto, the fastest way to lose money isn't choosing the wrong good project — it's buying something without having done any basic research at all. Using this framework, 1–2 hours of research per new project is the lowest-cost insurance against the recurring story of 'followed a Twitter recommendation, bought at peak, sold down 90%.' One additional reminder: research frameworks help you analyze 'known information,' but crypto markets still contain large volumes of 'unknown unknowns' — black swan events (hacks, regulatory crackdowns, founder exits) are things no amount of due diligence can fully predict. Position control (never allocating more than 5–10% of total capital to a single asset) is always your final safety net.