Traditional financial analysis relies heavily on data companies choose to disclose — quarterly filings, investor calls, occasional leaks. Public blockchains flip that model entirely: essentially every transaction is visible in real time to anyone willing to look, permanently. On-chain analysis is the practice of turning that raw transaction data into an actual read on market conditions, rather than treating it as an undifferentiated wall of numbers.

The building blocks

Before the metrics, it helps to know what's actually being measured.

  • Addresses — the public identifiers that send and receive assets. A single person can control many addresses, and many addresses (like exchange wallets) are controlled by one entity on behalf of thousands of users — a constant source of misreading if ignored.
  • Transactions — individual transfers or contract interactions, each permanently timestamped and traceable.
  • Blocks — batches of transactions, each referencing the one before it, which is what lets analysts reconstruct exact history and exact timing.
  • Smart contract events — logs emitted by contract interactions, which is how analysts track activity inside DeFi protocols specifically, not just simple transfers.

Metrics analysts actually watch

Exchange flows

The net movement of an asset onto or off of exchange addresses. Large inflows to exchanges are often read as a precursor to potential selling, since assets typically need to be on an exchange to be sold for another asset or currency. Large outflows are read the opposite way — as accumulation, since holders moving assets into self-custody are generally not preparing to sell imminently. Neither reading is certain; exchange flows are a probabilistic signal, not a fact about intent.

Holder distribution and cohort behavior

By clustering addresses into cohorts — by holding size, or by how long they've held without moving — analysts can see whether long-term holders are accumulating or distributing, separately from what short-term, more sentiment-driven participants are doing. Divergence between these cohorts (long-term holders accumulating while short-term activity is fearful, for instance) is one of the more closely watched patterns.

Realized value metrics

Rather than valuing every unit of an asset at today's price, these metrics value each unit at the price it last moved at, then sum across the network. This approximates the network's aggregate cost basis, and comparing it to current market value gives a rough sense of how much unrealized profit or loss the system as a whole is sitting on — a useful proxy for how much selling pressure might build if participants decide to lock in gains or cut losses.

Network activity

Active addresses, transaction counts and fees paid are basic but meaningful indicators of genuine usage, as distinct from pure price speculation. A rising price accompanied by falling network activity is a divergence worth noting; the two don't always move together for long.

Smart contract and protocol-level metrics

Within DeFi specifically, on-chain data reveals total value locked in a protocol, borrowing and lending activity, and liquidation levels — data that lets analysts see the health of a protocol's collateral base directly, rather than trusting a dashboard the protocol itself controls.

How this differs from technical chart analysis

Technical analysis studies price and volume patterns on a chart, largely agnostic to what's actually causing them. On-chain analysis studies the underlying activity generating those patterns — who's actually moving assets, how concentrated ownership is, and what's happening inside the protocols themselves. The two are complementary rather than competing: a price breakout that coincides with genuine growth in active addresses and exchange outflows tells a different story than the same breakout accompanied by falling activity and rising exchange inflows.

The caveats that matter most

On-chain data is complete but not automatically meaningful. The most common analytical mistake is treating a raw number as a signal without accounting for what's actually behind it.

  1. 01

    Address count isn't user count.

    One person can control thousands of addresses; one address (an exchange's) can represent millions of people. Metrics built on raw address counts need to be read with that in mind.

  2. 02

    Correlation across metrics isn't causation.

    Exchange inflows rising alongside a price drop doesn't prove the inflows caused the drop — both can be downstream of the same underlying sentiment shift.

  3. 03

    Off-chain activity is invisible.

    A large share of trading happens on centralized exchanges' internal order books, which never touch the public chain at all. On-chain data sees a real but partial picture.

  4. 04

    Sophisticated actors can obscure intent.

    Large holders routinely use multiple addresses, intermediary wallets and mixing techniques specifically to make their on-chain footprint harder to interpret.

Getting started

Block explorers and dedicated on-chain analytics platforms — the category covered in our toolkit — make this data browsable without writing code, with most offering free tiers sufficient for learning the basics. The most useful habit to build early is cross-referencing more than one metric before drawing a conclusion, since any single number in isolation is easy to misread.