Almost every explanation of blockchain starts with "a distributed, immutable ledger." That sentence is technically correct and completely useless if you don't already know what it means.
This guide takes a different route. We'll start with the actual problem blockchains were invented to solve, then build up the machinery piece by piece. By the end you'll understand not just how a blockchain works, but why each part exists — and, just as importantly, where the technology genuinely falls short.
The Problem Blockchains Solve
Digital Things Are Easy to Copy
Send someone a photo and you both have it. That's fine for photos. It's catastrophic for money.
If digital cash were just a file, you could send the same $10 file to a hundred people. This is called the double-spend problem, and until 2009 there was only one practical answer: put a trusted institution in the middle.
The Traditional Answer: A Trusted Middleman
Your bank keeps a ledger — a list of who owns what. When you send $50, the bank subtracts 50 from your row and adds 50 to someone else's. You can't double-spend because there's exactly one ledger and the bank controls it.
This works well, and for most people most of the time it's entirely adequate. But it has costs:
- You must trust the bank to keep honest records and stay solvent
- The bank can freeze or reverse your transactions
- The bank is a single point of failure — outages, hacks, or collapse affect everyone
- Access is permissioned — roughly 1.4 billion adults worldwide have no bank account at all
The Blockchain Answer: Everyone Keeps the Ledger
What if instead of one institution holding the ledger, thousands of independent computers each held an identical copy — and they had a reliable way to agree on updates without trusting each other?
That's a blockchain. The whole design exists to answer one question: how do strangers agree on a shared set of facts without a referee?
Everything below is machinery in service of that question.
The Building Blocks
Blocks
A block is a batch of transactions, bundled together with some metadata. Think of it as a page in a ledger.
Each block contains:
- A list of transactions ("Alice sent Bob 0.5 BTC")
- A timestamp
- A reference to the block before it
- A cryptographic fingerprint of its own contents
Hashes: The Cryptographic Fingerprint
A hash function takes any input and produces a fixed-length string of characters. Bitcoin uses SHA-256, which always outputs 64 hexadecimal characters.
Three properties make hashes useful here:
- Deterministic — the same input always produces the same output
- Avalanche effect — changing one character of the input changes the output completely
- One-way — you cannot work backwards from a hash to the original input
That second property is the important one. Hash the word "hello" and you get a specific 64-character string. Hash "Hello" — capital H — and you get an entirely different one, with no visible relationship to the first.
The Chain
Here's where "blockchain" gets its name. Every block includes the hash of the block before it.
That creates a dependency: Block 100 contains Block 99's hash. Block 99 contains Block 98's hash. And so on, back to the very first block.
Now consider what happens if someone tries to alter a transaction in Block 50:
- Changing it changes Block 50's contents
- Which changes Block 50's hash
- But Block 51 contains the old hash of Block 50 — so Block 51 is now invalid
- Fixing Block 51 changes its hash, invalidating Block 52
- ...and so on, all the way to the present
Tampering with one old block requires rewriting every block after it. This is what people mean by "immutable" — not that the data physically cannot change, but that changing it is prohibitively expensive and instantly detectable.
Consensus: How Strangers Agree
Chaining blocks makes tampering detectable. It doesn't yet stop someone from broadcasting a competing version of history. For that you need consensus — a rule for deciding which version everyone accepts.
Proof of Work
Bitcoin's approach. To add a block, participants called miners must find a number (a "nonce") which, when hashed together with the block's contents, produces a hash starting with a certain number of zeros.
There's no clever way to find it. You guess, hash, check, and repeat — billions of times per second. The network adjusts the difficulty so that, collectively, a solution is found roughly every ten minutes.
Why bother? Because it makes writing history expensive:
- Adding an honest block costs real electricity
- Rewriting history means redoing that work for every subsequent block, faster than the honest network is extending it
- To reliably rewrite Bitcoin's history you would need more computing power than the rest of the network combined — a 51% attack
The security guarantee isn't "this is impossible." It's "this costs more than you'd gain." That's an economic argument, not a mathematical one, and it's worth understanding the difference.
How miners are paid, and the halving. Each new Bitcoin block pays its miner a fixed number of newly created bitcoin (the block subsidy) plus the fees of the transactions inside it. The subsidy halves every 210,000 blocks, roughly every four years. The April 2024 halving cut it from 6.25 to 3.125 BTC per block; the next is expected around 2028. Over time, fees have to carry more of the cost of security. See What is Bitcoin? for the full supply schedule.
The trade-off: Proof of Work consumes serious energy — Bitcoin's network uses roughly as much electricity annually as a mid-sized country. Whether that cost is justified is a genuine debate, not a settled question.
Proof of Stake
Ethereum's approach since 2022, and the model most newer chains use. Instead of burning electricity, participants called validators lock up ("stake") the network's own currency as collateral — 32 ETH on Ethereum.
Validators are chosen to propose blocks roughly in proportion to their stake. If they behave dishonestly, the protocol destroys part of their stake — a penalty called slashing.
The security argument is the same shape: attacking the network requires acquiring an enormous stake, and a successful attack would devalue the very asset you had to buy. It's still "attacking costs more than it's worth," just with capital at risk rather than electricity.
The trade-off: Proof of Stake cuts energy use by well over 99%, but critics argue it favours those who already hold the most coins, and it's a younger design with less battle-testing than Proof of Work.
Why This Matters to You
You don't need to pick a side. But when you evaluate a crypto project, "how does this network reach consensus, and what does an attack cost?" is one of the few questions that genuinely separates serious projects from vapourware.
What Actually Happens When You Send Crypto
Let's trace a real transaction end to end.
1. You create and sign it. Your wallet builds a message: "move 0.1 BTC from this address to that address." It signs the message with your private key — a secret number only you hold.
2. Anyone can verify it, nobody can forge it. This is public-key cryptography. Your private key produces a signature; your matching public key lets anyone confirm the signature is genuine without ever revealing the private key. This is why "not your keys, not your coins" is repeated so often — the private key is the ownership.
3. It's broadcast. Your wallet sends the signed transaction to nearby nodes, which relay it onward. Within seconds it's known across the network and sitting in the mempool, a waiting area of unconfirmed transactions.
4. It gets included in a block. Miners or validators select transactions from the mempool — generally favouring those paying higher fees — and package them into a candidate block.
5. The block is added and propagated. Once the consensus rules are satisfied, the block is broadcast. Every node independently verifies it and appends it to their copy.
6. Confirmations accumulate. Your transaction now has one confirmation. Each subsequent block adds another. More confirmations mean more blocks would need rewriting to reverse it — which is why exchanges typically wait for several before crediting a deposit.
Public, Private, and the "Blockchain for Everything" Era
Around 2017–2018, seemingly every company announced a blockchain initiative. Most quietly disappeared. Understanding why is genuinely useful.
Public blockchains (Bitcoin, Ethereum) are open. Anyone can read them, submit transactions, or run a node. Nobody is in charge.
Private or permissioned blockchains restrict participation to approved parties. A consortium of banks might run one between themselves.
Here's the uncomfortable question for most private blockchain projects: if a small group of known, mutually accountable parties controls the network, what is the consensus mechanism actually protecting against? You've taken on the enormous performance cost of replicating a ledger across many machines, without the trustlessness that cost was buying. A well-designed shared database usually does the job better.
That's not to say private chains are always wrong — they have real uses in multi-party supply chains and settlement. But blockchain is a specific solution to the specific problem of coordinating between parties who don't trust each other. Applied anywhere else, it's expensive overhead.
Smart Contracts
Bitcoin's blockchain stores transactions. Ethereum generalised the idea: what if the chain could store and execute programs?
A smart contract is code deployed to the blockchain. Once deployed it runs exactly as written, automatically, and — critically — nobody can stop it or alter it, including its author.
A simple example: a contract holding funds that release to a seller only once a buyer confirms delivery. No escrow agent required; the code is the escrow agent.
This unlocked most of what people now call DeFi — lending protocols, decentralised exchanges, stablecoins. But "unstoppable code" cuts both ways:
- Bugs are permanent. If a contract has a flaw, it usually cannot be patched. Hundreds of millions of dollars have been lost to smart contract exploits.
- There is no support line. No chargebacks, no fraud department, no reversal.
- Audits reduce risk, they don't remove it. Plenty of audited contracts have been drained.
What Blockchains Are Genuinely Bad At
Any honest explanation has to include this section.
They're slow. Bitcoin handles about 7 transactions per second; Ethereum around 15–30. Visa handles tens of thousands. Making thousands of computers agree on every update is inherently slower than one database writing to disk.
They're expensive. During congestion, Ethereum fees have exceeded $50 for a single transaction. Since Ethereum's Dencun upgrade (March 2024), transactions on Layer 2 rollups usually cost cents, but the base layer will never be cheap by design at busy times.
They're bad at storing data. Every node stores everything forever. Putting a photo on-chain is wildly impractical. Blockchains store references and state, not files.
They can't verify the real world. A blockchain can prove a transaction occurred on its own ledger. It cannot prove a shipping container actually contains what a label claims. Anything entering from outside — via an "oracle" — reintroduces exactly the trust assumption you were trying to remove. This defeats a great many proposed use cases.
Privacy is weaker than people assume. Bitcoin is pseudonymous, not anonymous. Every transaction is permanently public. Chain-analysis firms routinely link addresses to real identities.
The scalability trilemma. It's broadly accepted that a blockchain struggles to be simultaneously decentralised, secure, and scalable. Chains boasting enormous throughput have almost always traded away decentralisation to get there. When you see a huge TPS number, ask what was given up.
Layer 2: Scaling Without Abandoning the Base
Since base layers are inherently constrained, most scaling work has moved to Layer 2 — systems that process transactions elsewhere and periodically settle to the main chain.
- Bitcoin's Lightning Network opens payment channels between parties. Two people can transact thousands of times instantly and near-free, settling only the final balance on-chain.
- Ethereum rollups (Arbitrum, Optimism, Base, zkSync) batch thousands of transactions off-chain and post compressed proofs to Ethereum, inheriting its security at a fraction of the cost.
The pattern: use the expensive, maximally secure base layer for final settlement, and do the high-volume work above it.
Key Concepts to Remember
- Block — a batch of transactions
- Hash — a fingerprint of data; any change produces a completely different one
- Node — a computer holding a copy of the chain and verifying rules
- Consensus mechanism — the rule for agreeing which version of history is real
- Private key — the secret that proves ownership; whoever holds it owns the funds
- Public key / address — what you share to receive funds
- Confirmation — a block added on top of yours, deepening its permanence
- Smart contract — code deployed on-chain that executes automatically
- Layer 2 — a system settling to a base chain to gain speed and cut cost
Key Takeaways
- Blockchains solve one specific problem: agreeing on shared facts without a trusted referee
- Chaining blocks by hash makes tampering detectable; consensus makes it expensive
- Security rests on economic arguments — attacking costs more than it yields — not mathematical impossibility
- Your private key is your ownership; nothing else can recover it
- The technology is genuinely poor at speed, cost, storage, privacy, and verifying real-world facts
- If the parties involved already trust each other, a database is almost certainly the better tool
- Understanding the trade-offs is what separates informed investing from following hype
Where to Go Next
Now that the machinery makes sense, the practical next steps are:
- Crypto Wallets Explained — securing the private keys that represent your ownership
- How to Buy Your First Cryptocurrency — putting this into practice safely
- DeFi Explained — what smart contracts made possible
- Crypto glossary — quick definitions when a term trips you up
Blockchain is neither the world-changing miracle its loudest advocates claim nor the pure scam its harshest critics insist on. It's a specific engineering trade-off: you give up speed, cost efficiency, and privacy, and you get coordination without a trusted intermediary. Whether that trade is worth making depends entirely on whether you actually needed to remove the intermediary.
Ask that question about every crypto project you evaluate. It will save you a great deal of money.