Output Linking

Output linking is the process by which blockchain analysts connect the outputs of one transaction to the inputs of later transactions and, ultimately, to a real-world identity. Bitcoin's UTXO model means every spend consumes prior outputs and creates new ones, leaving a permanent, public trail. On its own this data is pseudonymous noise — but the moment one address is tied to a known person (a KYC exchange withdrawal, a posted donation address, a merchant invoice), links propagate forward and backward through the chain, and a single identification can deanonymize a whole cluster of addresses and years of history.

How links are inferred

Analysts apply heuristics rather than certainties. The most powerful is the common-input-ownership heuristic: the assumption that all inputs to a transaction share one owner, which lets a single co-spend merge two address clusters into one. The second pillar is change detection — distinguishing the payment output from the change output that returns to the sender. Tells include round-number amounts (the 0.05 BTC output is probably the payment; the 0.03724... output is probably change), script types (change usually matches the input's address type), address freshness, and spending patterns over time. Address reuse is the analyst's best friend: every payment to a reused address enriches one well-labelled node in the graph. Each successful link tightens the cluster; commercial chain-analysis firms industrialize exactly this, layering exchange data and subpoenaed records on top of the public heuristics.

Breaking the links

Every practical Bitcoin privacy technique is, at its core, an attempt to make output linking unreliable. CoinJoin puts many users' inputs and equal-valued outputs in one transaction so the input-ownership assumption merges strangers and the output mapping becomes ambiguous. PayJoin inserts the recipient's input into an ordinary-looking payment, so the ownership heuristic yields a confidently wrong answer — quietly poisoning the analyst's model. CoinSwap removes the direct on-chain link between two of a user's positions entirely, and Silent Payments eliminate address reuse for donation-style receiving by deriving a fresh address per sender. The defender's goal is not perfect anonymity but degraded confidence: enough plausible deniability that any inferred link could be wrong, which in aggregate breaks the economics of mass surveillance.

Practical hygiene

For an ordinary self-custodial user, most linkage risk comes from lazy defaults, not exotic adversaries. Never reuse addresses. Use a wallet with coin control and label every UTXO when it arrives, so you know which coins carry which history before you co-spend them — the worst leaks come from casually combining a KYC-tagged coin with a private one in a single transaction, welding the two histories together forever. Think about consolidation the same way: sweeping everything into one output is cheap in fees and expensive in privacy. Miners should note that coinbase and pool-payout outputs are cleanly labelled at birth, making downstream linkage easier, not harder.

Output linking is the operational side of the transaction graph: the graph is the raw data, linking is the craft of reading it — and privacy practice is the craft of writing it illegibly. Self-custody gives you the pen; what the analysts can read is up to you.