This December I wrote a widely-circulated article on the inapplicability of blockchain to any actual problem. People objected mostly not to the technology argument, but rather hoped that decentralization could produce integrity.

Let’s start with this: Venmo is a free service to transfer dollars, and bitcoin transfers are not free. Yet after I wrote an article last December saying bitcoin had no use, someone responded that Venmo and Paypal are raking in consumers’ money and people should switch to bitcoin.

What a surreal contrast between blockchain’s non-usefulness/non-adoption and the conviction of its believers! It’s so entirely evident that this person didn’t become a bitcoin enthusiast because they were looking for a convenient, free way to transfer money from one person to another and discovered bitcoin. In fact, I would assert that there is no single person in existence who had a problem they wanted to solve, discovered that an available blockchain solution was the best way to solve it, and therefore became a blockchain enthusiast.

There is no single person in existence who had a problem they wanted to solve, discovered that an available blockchain solution was the best way to solve it, and therefore became a blockchain enthusiast.

The number of retailers accepting cryptocurrency as a form of payment is declining, and its biggest corporate boosters like IBM, NASDAQ, Fidelity, Swift and Walmart have gone long on press but short on actual rollout. Even the most prominent blockchain company, Ripple, doesn’t use blockchain in its product. You read that right: the company Ripple decided the best way to move money across international borders was to not use Ripples.

A blockchain is a literal technology, not a metaphor

Why all the enthusiasm for something so useless in practice?

People have made a number of implausible claims about the future of blockchain—like that you should use it for AI in place of the type of behavior-tracking that google and facebook do, for example. This is based on a misunderstanding of what a blockchain is. A blockchain isn’t an ethereal thing out there in the universe that you can “put” things into, it’s a specific data structure: a linear transaction log, typically replicated by computers whose owners (called miners) are rewarded for logging new transactions.

There are two things that are cool about this particular data structure. One is that a change in any block invalidates every block after it, which means that you can’t tamper with historical transactions. The second is that you only get rewarded if you’re working on the same chain as everyone else, so each participant has an incentive to go with the consensus.

The end result is a shared definitive historical record. And, what’s more, because consensus is formed by each person acting in their own interest, adding a false transaction or working from a different history just means you’re not getting paid and everyone else is. Following the rules is mathematically enforced—no government or police force need come in and tell you the transaction you’ve logged is false (or extort bribes or bully the participants). It’s a powerful idea.

So in summary, here’s what blockchain-the-technology is: “Let’s create a very long sequence of small files — each one containing a hash of the previous file, some new data, and the answer to a difficult math problem — and divide up some money every hour among anyone willing to certify and store those files for us on their computers.”

Now, here’s what blockchain-the-metaphor is: “What if everyone keeps their records in a tamper-proof repository not owned by anyone?”

An illustration of the difference: In 2006, Walmart launched a system to track its bananas and mangoes from field to store. In 2009 they abandoned it because of logistical problems getting everyone to enter the data, and in 2017 they re-launched it (to much fanfare) on blockchain. If someone comes to you with “the mango-pickers don’t like doing data entry,” “I know: let’s create a very long sequence of small files, each one containing a hash of the previous file” is a nonsense answer, but “What if everyone keeps their records in a tamper-proof repository not owned by anyone?” at least addresses the right question!

Blockchain-based trustworthiness falls apart in practice

People treat blockchain as a “futuristic integrity wand”—wave a blockchain at the problem, and suddenly your data will be valid. For almost anything people want to be valid, blockchain has been proposed as a solution.

It’s true that tampering with data stored on a blockchain is hard, but it’s false that blockchain is a good way to create data that has integrity.

It’s true that tampering with data stored on a blockchain is hard, but it’s false that blockchain is a good way to create data that has integrity.

To understand why this is the case, let’s work from the practical to the theoretical. For example, let’s consider a widely-proposed use case for blockchain: buying an e-book with a “smart” contract. The goal of the blockchain is, you don’t trust an e-book vendor and they don’t trust you (because you’re just two individuals on the internet), but, because it’s on blockchain, you’ll be able to trust the transaction.

In the traditional system, once you pay you’re hoping you’ll receive the book, but once the vendor has your money they don’t have any incentive to deliver. You’re relying on Visa or Amazon or the government to make things fair—what a recipe for being a chump! In contrast, on a blockchain system, by executing the transaction as a record in a tamper-proof repository not owned by anyone, the transfer of money and digital product is automatic, atomic, and direct, with no middleman needed to arbitrate the transaction, dictate terms, and take a fat cut on the way. Isn’t that better for everybody?

Hm. Perhaps you are very skilled at writing software. When the novelist proposes the smart contract, you take an hour or two to make sure that the contract will withdraw only an amount of money equal to the agreed-upon price, and that the book — rather than some other file, or nothing at all — will actually arrive.

Auditing software is hard! The most-heavily scrutinized smart contract in history had a small bug that nobody noticed — that is, until someone did notice it, and used it to steal fifty million dollars. If cryptocurrency enthusiasts putting together a $150m investment fund can’t properly audit the software, how confident are you in your e-book audit? Perhaps you would rather write your own counteroffer software contract, in case this e-book author has hidden a recursion bug in their version to drain your ethereum wallet of all your life savings?

It’s a complicated way to buy a book! It’s not trustless, you’re trusting in the software (and your ability to defend yourself in a software-driven world), instead of trusting other people.

Another example: the purported advantages for a voting system in a weakly-governed country. “Keep your voting records in a tamper-proof repository not owned by anyone” sounds right — yet is your Afghan villager going to download the blockchain from a broadcast node and decrypt the Merkle root from his Linux command line to independently verify that his vote has been counted? Or will he rely on the mobile app of a trusted third party — like the nonprofit or open-source consortium administering the election or providing the software?

These sound like stupid examples — novelists and villagers hiring e-bodyguard hackers to protect them from malicious customers and nonprofits whose clever smart-contracts might steal their money and votes?? — until you realize that’s actually the point. Instead of relying on trust or regulation, in the blockchain world, individuals are on-purpose responsible for their own security precautions. And if the software they use is malicious or buggy, they should have read the software more carefully.

The entire worldview underlying blockchain is wrong

You actually see it over and over again. Blockchain systems are supposed to be more trustworthy, but in fact they are the least trustworthy systems in the world. Today, in less than a decade, three successive top bitcoin exchanges have been hacked, another is accused of insider trading, the demonstration-project DAO smart contract got drained, crypto price swings are ten times those of the world’s most mismanaged currencies, and bitcoin, the “killer app” of crypto transparency, is almost certainly artificially propped up by fake transactions involving billions of literally imaginary dollars.

Blockchain systems do not magically make the data in them accurate or the people entering the data trustworthy, they merely enable you to audit whether it has been tampered with. A person who sprayed pesticides on a mango can still enter onto a blockchain system that the mangoes were organic. A corrupt government can create a blockchain system to count the votes and just allocate an extra million addresses to their cronies. An investment fund whose charter is written in software can still misallocate funds.

How then, is trust created?

In the case of buying an e-book, even if you’re buying it with a smart contract, instead of auditing the software you’ll rely on one of four things, each of them characteristics of the “old way”: either the author of the smart contract is someone you know of and trust, the seller of the e-book has a reputation to uphold, you or friends of yours have bought e-books from this seller in the past successfully, or you’re just willing to hope that this person will deal fairly. In each case, even if the transaction is effectuated via a smart contract, in practice you’re relying on trust of a counterparty or middleman, not your self-protective right to audit the software, each man an island unto himself. The contract still works, but the fact that the promise is written in auditable software rather than government-enforced English makes it less transparent, not more transparent.

The same for the vote counting. Before blockchain can even get involved, you need to trust that voter registration is done fairly, that ballots are given only to eligible voters, that the votes are made anonymously rather than bought or intimidated, that the vote displayed by the balloting system is the same as the vote recorded, and that no extra votes are given to the political cronies to cast. Blockchain makes none of these problems easier and many of them harder—but more importantly, solving them in a blockchain context requires a set of awkward workarounds that undermine the core premise. So we know the entries are valid, let’s allow only trusted nonprofits to make entries—and you’re back at the good old “classic” ledger. In fact, if you look at any blockchain solution, inevitably you’ll find an awkward workaround to re-create trusted parties in a trustless world.

A crypto-medieval system

Yet absent these “old way” factors—supposing you actually attempted to rely on blockchain’s self-interest/self-protection to build a real system—you’d be in a real mess.

Eight hundred years ago in Europe — with weak governments unable to enforce laws and trusted counterparties few, fragile and far between — theft was rampant, safe banking was a fantasy, and personal security was at the point of the sword. This is what Somalia looks like now, and also, what it looks like to transact on the blockchain in the ideal scenario.

Somalia on purpose. That’s the vision. Nobody wants it!

Even the most die-hard crypto enthusiasts prefer in practice to rely on trust rather than their own crypto-medieval systems. 93% of bitcoins are mined by managed consortiums, yet none of the consortiums use smart contracts to manage payouts. Instead, they promise things like a “long history of stable and accurate payouts.” Sounds like a trustworthy middleman!

Same with Silk Road, a cryptocurrency-driven online drug bazaar. The key to Silk Road wasn’t the bitcoins (that was just to evade government detection), it was the reputation scores that allowed people to trust criminals. And the reputation scores weren’t tracked on a tamper-proof blockchain, they were tracked by a trusted middleman!

If Ripple, Silk Road, Slush Pool, and the DAO all prefer “old way” systems of creating and enforcing trust, it’s no wonder that the outside world had not adopted trustless systems either!

In the name of all blockchain stands for, it’s time to abandon blockchain

A decentralized, tamper-proof repository sounds like a great way to audit where your mango comes from, how fresh it is, and whether it has been sprayed with pesticides or not. But actually, laws on food labeling, nonprofit or government inspectors, an independent, trusted free press, empowered workers who trust whistleblower protections, credible grocery stores, your local nonprofit farmer’s market, and so on, do a way better job. People who actually care about food safety do not adopt blockchain because trusted is better than trustless. Blockchain’s technology mess exposes its metaphor mess — a software engineer pointing out that storing the data a sequence of small hashed files won’t get the mango-pickers to accurately report whether they sprayed pesticides is also pointing out why peer-to-peer interaction with no regulations, norms, middlemen, or trusted parties is actually a bad way to empower people.

Like the farmer’s market or the organic labeling standard, so many real ideas are hiding in plain sight. Do you wish there was a type of financial institution that was secure and well-regulated in all the traditional ways, but also has the integrity of being people-powered? A credit union’s members elect its directors, and the transaction-processing revenue is divided up among the members. Move your money! Prefer a deflationary monetary policy? Central bankers are appointed by elected leaders. Want to make elections more secure and democratic? Help write open source voting software, go out and register voters, or volunteer as an election observer here or abroad! Wish there was a trusted e-book delivery service that charged lower transaction fees and distributed more of the earnings to the authors? You can already consider stated payout rates when you buy music or books, buy directly from the authors, or start your own e-book site that’s even better than what’s out there!

Projects based on the elimination of trust have failed to capture customers’ interest because trust is actually so damn valuable. A lawless and mistrustful world where self-interest is the only principle and paranoia is the only source of safety is a not a paradise but a crypto-medieval hellhole.

As a society, and as technologists and entrepreneurs in particular, we’re going to have to get good at cooperating — at building trust, and, at being trustworthy. Instead of directing resources to the elimination of trust, we should direct our resources to the creation of trust—whether we use a long series of sequentially hashed files as our storage medium or not.

Kai Stinchcombe coined the terms “crypto-medieval” “futuristic integrity wand” and “smart mango.” Please use freely: coining terms makes you a futurist.

Everyone says the blockchain, the technology underpinning cryptocurrencies such as bitcoin, is going to change EVERYTHING. And yet, after years of tireless effort and billions of dollars invested, nobody has actually come up with a use for the blockchain—besides currency speculation and illegal transactions.

Each purported use case — from payments to legal documents, from escrow to voting systems—amounts to a set of contortions to add a distributed, encrypted, anonymous ledger where none was needed. What if there isn’t actually any use for a distributed ledger at all? What if, ten years after it was invented, the reason nobody has adopted a distributed ledger at scale is because nobody wants it?

Payments and banking

The original intended use of the blockchain was to power currencies like bitcoin — a way to store and exchange value much like any other currency. Visa and MasterCard were dinosaurs, everyone proclaimed, because there was now a costless, instant way to exchange value without the middleman taking a cut. A revolution in banking was just the start… governments, unable to issue currency by fiat anymore, would take a back seat as individual citizens transacted freely outside any national system.

It didn’t take long for that dream to fall apart. For one thing, there’s already a costless, instant way to exchange value without a middleman: cash. Bitcoins substitute for dollars, but Visa and MasterCard actually sit on top of dollar-based banking transactions, providing a set of value-added services like enabling banks to track fraud disputes, and verifying the identity of the buyer and seller. It turns out that for the person paying for a product, the key feature of a new payment system — think of PayPal in its early days — is the confidence that if the goods aren’t as described you’ll get your money back. And for the person accepting payment, basically the key feature is that their customer has it, and is willing to use it. Add in points, credit lines, and a free checked bag on any United flight and you have something that consumers choose and merchants accept. Nobody actually wants to pay with bitcoin, which is why it hasn’t taken off.

The key feature of a new payment system — think of PayPal in its early days — is the confidence that if the goods aren’t as described you’ll get your money back.

Plus, it’s not actually that good a payment system — Visa can handle sixty thousand transactions per second, while Bitcoin historically taps out at seven. There are technical modifications going on to improve Bitcoin’s efficiency, but as a starting point, you have something that’s about 0.01% as good at clearing transactions. (And, worth noting, for those seven transactions a second Bitcoin is already estimated to use 35 times as much energy as Visa. If you brought Bitcoin’s transaction volume up to Visa’s it would be using as much electricity as the rest of the world put together.)

Freedom to transact without government supervision

In many countries, and often our own, a little bit of ability to keep a few things private from the authorities probably makes the world a better place. In places like Cuba or Venezuela, many prefer to transact in dollars, and bitcoin could in theory serve a similar function. Yet there are two reasons this hasn’t been the panacea it’s assumed: the advantages of government to the individual, and the advantages of government to society.

The government-backed banking system provides FDIC guarantees, reversibility of ACH, identity verification, audit standards, and an investigation system when things go wrong. Bitcoin, by design, has none of these things. I saw a remarkable message thread by someone whose bitcoin account got drained because their email had been hacked and their password was stolen. They were stunned to have no recourse! And this is widespread — in 2014, the then-#1 bitcoin trader, Mt. Gox, also lost $400m of investor money due to security failures. The subsequent #1 bitcoin trader, Bitfinex, also shut down after a loss of customer funds. Imagine the world if more banks had been drained of customer funds than not. Bitcoin is what banking looked like in the middle ages — “here’s your libertarian paradise, have a nice day.”

Bitcoin is what banking looked like in the middle ages — “here’s your libertarian paradise, have a nice day.

[This issue is particularly near and dear to my heart because my own company, True Link, is designed to help vulnerable seniors — people likely to give out their credit card number over the phone, enter sketchy sweepstakes or donate to sketchy charities, participate in scam investments, or install password-stealing malware. As the people who most need security enhancements in banking and payments, they depend heavily on the existing protections and would absolutely be harmed by many of the proposed changes in favor of private-key authenticated, instant, and irreversible transfers. Someone starting from a human perspective on banking security—who is currently harmed and how can we help them?—would come up with something very different from blockchain!]

Second, government policies are designed to disrupt terrorist financing and organized crime, and prevent traffic in illegal goods like stolen credit card numbers or child pornography. The mainstream preference is to have transactions private but not undiscoverable under warrant — ask “should the government have a list everyone you’ve paid money to,” and most will say no; ask “should the government be able under warrant to get a list everyone a child pornography collector has paid money to,” and most will say yes. Nobody wants bitcoin to 100x the total traffic in goods and services our government defines as illegal — as one bitcoin enthusiast pointed out to me, “If you invented cash today, it would be illegal too.”

Micropayments and bank-to-bank transfers

It’s worth noting two particular payment use cases where people are particularly excited about blockchain-based currencies: micropayments and bank-to-bank transfers. In terms of micropayments, people enthuse that bitcoin transactions are free and instant. Actually, they take about eight minutes to clear and cost about four cents to process. People have proposed that you will use bitcoins for micropayments — for example, paying two cents to a musician to listen to their song on the internet, or four cents to read a newspaper article. Yet the infrastructure to do this — for example, advance authorization with the source of funds so you don’t have to wait eight minutes to read the article you just clicked — actually eliminates the need for bitcoin at all. If you’re happy to pay four cents an article or two cents a song, you can set it up to bill once a month from your bank account and read to your heart’s content. And in practice, people prefer subscription services to micropayments.

In terms of interbank payments, many people mention Ripple as a promising way to transfer money between banks. Over the last 30 days it processed two billion dollars (as of this writing) worth of interbank and interpersonal transactions — about 40 seconds’ worth of volume on the SWIFT interbank network — after three years of being available to banks to trade 90% of the world’s high-volume currencies. This is like the proportion of US GDP comprised by toothpick sales. Why haven’t banks preferred this new technology? The answer is that setting up a Ripple Gateway isn’t actually much different than using the existing corresponding-account system — except that a lost password or security token can lead to much larger and more instant actual losses — which, as a reminder, has happened to more leading bitcoin exchanges than have managed to avoid it. The same features that make the banking system attractive to end users also make it attractive to banks. They already have ledgers, and don’t need to distribute them, anonymize them, encrypt them, publish them, and make them irreversible.

“Smart” contracts

“Smart” contracts are contracts written as software, rather than written as legal text. Because you can encode them directly on the blockchain, they can involve the transfer of value based directly on the cryptographic consent of the parties involved — in other words, they are “self-executing.” And in theory, contracts written in software are cheaper to interpret — because their operation is literally mathematical and automatic, there are no two ways to interpret them, which means there’s no need for expensive legal battles.

And yet the real-world examples show the ways this is problematic. The most prominent and largest smart contract to date, an investment vehicle called the Distributed Autonomous Organization (DAO), enabled its members to invest directly using their private cryptographic keys to vote on what to invest in. No lawyers, no management fees, no opaque boardrooms, the DAO “removes the ability of directors and fund managers to misdirect and waste investor funds.” And yet, due to a software bug, the DAO “voted” to “invest” $50m, a third of its members’ money, into a vehicle controlled by very clever programmers who knew a lot about recursion issues during balance updates. Some said this was a hack or an exploit because the software had not functioned as intended, while others said that there was no such thing as a hack — the whole point was that the software made decisions autonomously and there were no two ways to interpret it, and if you didn’t understand how the software worked you shouldn’t have participated. In the end, everyone got together and voted to retroactively amend the software contract and move the money back to its original owners. What’s the takeaway? Even the most die-hard blockchain enthusiasts actually want a bunch of humans arguing about the underlying intention behind a contract, rather than letting the software self-execute. Maybe the “dumb” way is smart after all?

The DAO was an illustrative experiment, but what about for routine transactions at big companies? The investors and startups in the smart-contract space promise that the block chain will enable super-fast execution and payment — for example that in healthcare applications, “instead of waiting 90–180 days for a claim to be processed, or spending hours on the phone trying to get your bill paid, it can in theory be processed on the spot.” But that’s true for any software-enabled purchasing system. My company’s Amazon servers scale automatically based on website traffic and bill us for how much we use. The idea that smart contracts would change this is a fallacy — it conflates the legal arrangement being put into effect with software with the legal arrangement itself being coded as software. Amazon’s terms of service are not a smart contract, but the billing system that implements those terms is automated. To the extent that health insurance billing, for example, is not automated, the problem isn’t that existing software isn’t “smart” enough to handle submitting claims and paying them electronically, it’s that the insurance company is slow moving, either by accident or because they on-purpose prefer a human review.

In the end, everyone from blockchain enthusiasts to health insurers actually wants to argue out in human language what the business relationship is and interpret it on an ongoing basis, and then to write software that handles the fulfillment and payment. That already exists — it’s the status quo.

Distributed storage, computing, and messaging

Another implausible idea is using the blockchain as a distributed storage mechanism. On its face it makes sense — you break your document up into “blocks”, encrypt them, and put them in a distributed ledger… it’s backed up across multiple locations, it’s secure, and easy to track everything that happened.

Yet there are multiple excellent ways to break up files, encrypt them, and replicate them across multiple storage media in different locations. There is already a company that bills itself as a cheaper, distributed Dropbox, which encrypts and stores files across multiple users’ hard drives and pays them a small fee for the free space on their hard drives. The block chain is just a particularly inefficient and insecure way of doing this.

There are four additional problems with a blockchain-driven approach. First, you’re relying on single-point encryption — your own private keys — rather than a more sophisticated system that might involve two-factor authorization, intrusion detection, volume limits, firewalls, remote IP tracking, and the ability to disconnect the system in an emergency. Second, price tradeoffs are entirely implausible — the bitcoin blockchain has consumed almost a billion dollars worth of electricity to hash an amount of data equivalent to about a sixth of what I get for my ten dollar a month dropbox subscription. Fourth, systematically choosing where and how much to replicate data is an advantage in the long run — the blockchain’s defaults on data replication just aren’t that smart. And finally, Dropbox and Box.com and Google and Microsoft and Apple and Amazon and everyone else provide a set of valuable other features that you don’t actually want to go develop on your own. Analogous to Visa, the problem isn’t storing data, it’s managing permissions, un-sharing what you shared before, getting an easy-to-view document history, syncing it on multiple devices, and so on.

The same argument holds for proposed distributed computing and secure messaging applications. Encrypting it, storing it forever, and replicating it across the entire network is just a ton of overhead relative to what you’re actually trying to accomplish. There are excellent computing, messaging, and storage solutions out there that have all the encryption and replication anyone needs — actually better than blockchain based solutions — and have plenty of other great features in addition.

Stock issuance

It was much-heralded when NASDAQ launched an internal blockchain-driven exchange for privately-held stocks. But wait: correct me if I’m wrong, but the whole purpose of NASDAQ (or the DTCC trade clearing system, for example) is that it has a ledger of who owns what stocks? Were they nervous that their systems, absent blockchain, would soon be unable to keep track of who owns what?

Similar to other transaction-tracking problems such as customer-to-merchant payments, the difference between NASDAQ’s ledger and blockchain’s ledger is that blockchain is distributed — it addresses the problem of lack of a trusted intermediary. And yet (for legal transactions) the company itself, its transfer agent of record, a clearinghouse, or an exchange are all trusted intermediaries and typically provide value-added services in addition. The reason NASDAQ is the right home for a blockchain-driven exchange is that they’re expert in the compliance and security aspects of trading stock. Cut out the middleman (here, NASDAQ itself) and the government and you’ll ultimately be limited to companies that choose to make an end-run around the legal, compliance, and tracking systems common to the mainstream market. As people who trade in unlisted stocks will tell you, that’s a recipe for getting your money stolen.

And we’re already seeing this. New companies have also begun creating blockchain-based “coins” convertible into company stock, and selling them to the public in Initial Coin Offerings, or ICOs, as a cheaper and more flexible way to raise money than a traditional Initial Public Offering of stocks on an exchange. It will be interesting to see how long this craze lasts — among other things, offering tokens convertible to stock counts as a securities offering, and so the SEC rules presumably apply to these securities offerings just like any other. Either the “coins” are just less-secure electronic stock certificates — protected by however carefully you store your password, rather than by the laws and protections of a securities exchange — or it’s another attempt to do an end-run around the law.

Authenticity verification

Another plausible use of the blockchain is that if you want to make a public, unalterable, undeleteable signed statement, you can “publish” it to the block chain — thinking of the distributed ledger as more like a diary than a way to buy and sell. In theory you could use this for recording vote tallies, verifying the origin of diamonds or brand-name gear, verifying people’s identity, resolving the ownership of domain names, keeping items in escrow, disclosing provisional patents under seal, notarizing documents, and so on.

Without diving too thoroughly into the details of each of these, it seems the use cases all fall apart pretty quickly. For voting, the status quo is recording the total number of ballots cast, with the voter dropping a visible paper ballot in a box, and journalists and observers from both sides watching the ballot boxes the whole time. The tough problem in voting is keeping who voted for who anonymous and yet making sure that voters and votes are one to one. Paper does this so much better than blockchain.

For a public notary or similar, verifying your driver’s license or having witnesses known to you present means that it wasn’t signed with a stolen password or private key — but, if a password or private key is adequate, you can just publish it signed with a PGP key. For establishing the authenticity of brand name goods like watches or handbags, or that a diamond was ethically mined, the ledger being distributed and encrypted doesn’t add any value — the originating company can just include a certificate you can verify online, just as they have done in the past. In cases of escrow, a smart contract can automatically pay for the goods without a need for a third party to verify and hold the funds, but you still need a trusted party to verify that the goods are delivered and as-promised.

And finally, if you want to irrefutably prove that you knew X at time Y without disclosing the actual knowledge publicly, encrypt it and email it to yourself at both a gmail and a hotmail address or post it on bitbucket, or print it out and notarize it, or postmark it by mailing it to yourself, or tweet an md5 of it, or whatever. But then again, how large is the irrefutably-prove-you-knew-X-at-time-Y-without-disclosing-X industry? Can you think of any leading company, or any company at all, that provides this service?

For domain resolution — the process of figuring out whose servers get to see the traffic and respond to your requests when you type a URL into your address bar — it’s promising to imagine that an all-digital record of smart contracts, where the actual act of payment being published to the ledger also updates who the domain resolves to, obviating the need for domain escrow services. Yet in practice, as with the DAO or other smart contracts, if valuable domains change hands due to theft or security issues, you actually need a way to override the ledger — as the result of a court order, for example. Just like with government-backed, law-backed bank accounts, real companies won’t prefer a situation in which a security breach or stolen password could result in someone else permanently and irrevocably owning bankofamerica.com or disney.com or sony.com or whatever. Adopting block chain technology makes theft or impersonation more likely rather than less. It sounds hypothetical until you realize more leading bitcoin exchanges have been hacked than not — something that very rarely happens with the leading domain name providers.

So what’s left?

Each of these seems trivial — yes, everyone knows handbags already come with certificates of authenticity with an ID number you can look up online — except that in each case, millions if not tens of millions of dollars have been spent on entire companies dedicated to just that particular use case. And you can get even more esoteric — Second Life on the blockchain, or blockchain-enabled appliances so your washing machine can smart-contract for its own detergent, or a sports league where the coaching decisions are written on the blockchain. (For real!)

In the end, the advantages of the existing human and software systems surrounding transactions — from verifying identity with a driver’s license to calling and clarifying the statements made in a credit disputed transaction to automatically billing your credit card for a newspaper subscription — outweigh the purported benefits, as well as hidden costs, of irrevocable, automated execution. Blockchain enthusiasts often act as if the hard part is getting money from A to B or keeping a record of what happened. In each case, moving money and recording the transaction is actually the cheap, easy, highly-automated part of a much more complex system.

Nobody went out and did a survey about whether most credit card users would be willing to give up their frequent flyer miles in return for also losing the ability to dispute a transaction.

Which leaves us where we started — currency speculation and illegal transactions — along with perhaps a lesson. In conversations with bitcoin entrepreneurs and investors and consultants, there was often a lack of knowledge or even interest in how the jobs were being done today or what the value to the end user was. With all the money spent on bitcoin cash registers, nobody went out and did a survey about whether most credit card users would be willing to give up their frequent flyer miles in return for also losing the ability to dispute a transaction. Presumably, they thought, the reason IPOs are so expensive or venture fund formation paperwork is so onerous is because all those lawyers and accountants are just getting rich sitting around pushing paper… a bunch of smart engineers in their 20s with no industry experience could certainly do their jobs, automatically, in a matter of months, with just a few million bucks of venture capital.

So far, not so much.

Kai Stinchcombe is CEO and cofounder of True Link Financial, a banking and investment service for seniors. In his spare time he enjoys hoping that, post singularity, a detergent delivery drone doesn’t self-execute a smart contract on his life, bitbleaching him from the sky into a hissing pool of unstructured data in exchange for a handful of bitcoins.

Five years ago, my co-founder Claire McDonnell and I founded True Link Financial, and did two things that that were totally unorthodox for the time.

First, we didn’t build a startup for teenagers.

We built it for seniors. (We’re a financial services firm that offers debit and Visa cards and investments management for aging Americans.)

Secondly, we charged our customers actual, real money for what we had built.

This was not common practice at the time. Around 2012, we and a whole cohort of other startups — Simple, Plastc, Swyp, Coin, Final, Stratos, Clinkle, and others — were getting card-issuing companies off the ground.¹ We also launched a Visa card — but our goal was to protect older folks from fraud, a $36 billion problem that affects millions of Americans, including my grandmother. Our cards would automatically decline scammy transactions so seniors like my grandmother would be able to still carry a credit card, preserve their independence, and spend their own money — and in order to decline the transactions we had to become the issuer of the card.²

The other companies offered a no-fees, high-end products with attractive features like extra security for online purchases, easier access to customer support, or perks and rewards. With expensive technology and support costs, we knew the companies offering free cards were losing money on every card they issued.³ In contrast, we charged $10 a month for each card.

With splashy videos and tons of press, some of our peers had waiting lists of 50,000 customers or more by the time they launched.

That wasn’t us.

When we launched our product we probably had a waitlist of 50. If you took me, my roommate, my mom, my dad, my co-founder, and my best friend out, it was more like 44. By our first year in market it had grown to about 1,000 customers — about 2 percent of where competitors got to before they had even launched. The two VC funds we wanted to work with the most said, “True Link would be so great if you just knew how to get customers,” and even better, “So, how come nobody wants it?”

Garry Tan, one of our first investors, sent Ash Rust over to our office to deliver a message: “You need to raise more money now, not later.” He thought that our strategy was going to be a long slog and that we didn’t have room to cut it close.

Charging for your product instills discipline

We made the case to our investors and potential investors that what we had done was actually more difficult and more impressive. Just ask the promoters handing out energy drink samples at the BART entrances: getting 50 people to accept something for free is easier than getting one person to pay $10 for it. If the other companies asked everyone on their waitlist to pay for their product, how many would stick with it? We argued that our 1,000 customers were better signal than the other companies’ 50,000 customers.

“We didn’t want to teach ourselves how to find money-losing customers.”

Right or wrong about the signal, what was more important was the discipline it created internally.

We didn’t want to teach ourselves how to find money-losing customers. We wanted to teach ourselves how to find customers that would sustain the business. We knew it would be harder and we would have to try more things. But while huge backlogs of money-losing customers might be good for raising capital, ultimately we wanted to build a product worth purchasing — and, a business that could last. Issuing and servicing cards actually costs money, and it doesn’t help either us or our customers to sign them up for a money-losing business and then disappear.

We found that what other card startups were doing — press, Facebook ads, and punchy video content — didn’t generate paid signups. It was enough to generate weak interest — “Sure, I’d try that” — from early adopters, but not enough to generate strong interest — “I’ve decided to close my other credit cards and use this instead.”

The people with the massive waitlists, it turns out, hadn’t really sold their product yet, they just convinced people to take a free trial. And not only does this not build a sustainable business, but the impact we were looking to have — protecting seniors from billions of dollars of fraud — required people to close their other credit cards and use ours instead.

In contrast, we found that word of mouth convinced people to actually buy and use the product. If someone you trust says it worked for them or someone they know, you’re ready to believe it’s a real solution, worth paying for. Everyone wants viral growth, but building for referrals is a very different thing — and it sometimes means foregoing paid channels that produce quicker results.

Without charging, we would have built the wrong company

Today, 95 percent of our cards and investment products are set up as a result of an in-person conversation — either with a family member, friend, or trusted professional like an in-home care provider — who already has experience with the product. And remarkably, in 85 percent of those cases we know who helped the customer sign up and how that person heard about us. We can trace dollars of revenue, for example, back to conferences we presented at to find experts in the aging market, and how those influencers spread the word.

Everything internally looks different as a result. Because our existing customers are our marketing engine, we’ve found from internal surveys that our customer support has a 99 percent satisfaction rating. Our customers are the primary way that we grow.

We’ve built software to make it easy for different types of key players to start the on-boarding process. And our marketing relies on a detailed set of referral analytics tools that few other card issuers have.

Suppose, early on, our twenty word-of-mouth customers had been thrown in with a thousand free customers coming in off a good Facebook video ad. Would we have built marketing tools to find more word-of-mouth customers?

The point is, if we’d built for an audience of free customers rather than an audience of paying customers, it would have been a different marketing strategy and team, different technology, different support, different analytics, different everything. We would have built the wrong company, and when it came time to make money we would be starting from square one.

Instead, we built the right business

We always believed that the marketing experiments we were running were tougher but more valuable. There is no dodging the fact that it sucked being compared to companies with a waiting list a thousand times longer. Some days, the early successes felt like tiny sparks, but over time we found we could turn the sparks into fires. Most things that could generate unpaid sign-ups failed to generate paid sign-ups, but we doubled down on the exceptions and iterated rapidly, quickly setting aside what wasn’t working.

We’d planned all along to do something different and more difficult than everyone else. Matt Karls, our board member from healthcare-focused venture fund Cambia, told his coworkers about how we put lines of tape across on the floor every month to show how much of our rent we were paying out of revenue and how fast that was growing. To us, it was obvious — companies that can’t pay their rent cease to exist. And at True Link, we wanted to keep existing.

Our cards division hit profitability earlier this year. Just three years after we were getting beat up for not having the traction our peers had, we were the first among them to break even — and by a huge margin. If we’d been selling a thousand free cards a week, we probably would never have noticed the twenty customers that actually cared enough to pay for our product — the customers who ended up making True Link a sustainable, long-term business.

True Link Financial provides Visa cards and investment portfolios for the aging population with a focus on financial protection. The company announced its Series A financing on October 11.

— — — — — — —

[1] Note that some were issuers and others were wrappers around other issuers, and some mixed a little of both or were using one as a launchpad into the other. We’re friends with people at Simple, Coin, Final, and Plastc. Each of them — and likely the others we don’t know — are great entrepreneurs, we shared our lessons both ways freely and openly, and hopefully they learned from us just like we learned from them. Simple got an awesome acquisition and is still growing, Final has just finished launching a credit product as a startup which is an incredible feat, and Coin and Plastc did amazing hardware innovations and each gave us really valuable advice. I’d love to see each of them write a blog post like this one about the times they got kicked in the teeth and bounced back — their stories are at least as good as ours and in some cases better. Nothing in here is intended as a knock against anyone. Except for Clinkle, which, I mean….

[2] An important technical distinction: these startups do not issue the card, but rather manage the card issuing program.

[3] For context, credit cards make money in four ways: monthly or annual fees, usage fees (like at ATMs, or for overdrafts), interest, and what’s called “interchange” — when you swipe your credit card, merchants get 97¢ on the dollar and interchange is the one cent that the card issuer gets. Mature card issuers carefully model their revenues to exceed their costs on average at scale — for example, for high-end cards, points and miles almost equal revenue from interchange, credit defaults almost equal the revenue from interest, and the issuer gets a tiny slice of both and makes money on the annual fee. If you omit an annual or monthly fee, you’re banking on a slim, slim slice of interchange adding up at scale, and if you are a debit card (no interest) and don’t charge overdraft fees (for example) you are betting on a miracle level of scale, as well as patient and plentiful capital to get there.

How charging for our product originally hurt — but ultimately saved — our startup was originally published in Initialized Capital on Medium, where people are continuing the conversation by highlighting and responding to this story.

Today’s venture capitalists can’t get through a day without getting pitched on how AI and chatbots are revolutionizing customer experience. See below for the the brave new world we can come to expect from our banks, our phone carriers, our healthcare providers, and eventually, God forbid, the government. [[Coda: I ended up getting a LOVELY note from a real human — a note that was everything I’d hoped for.]]

Me, via the “report an issue” button: My driver made an unplanned stop. Had to drop something off for a previous passenger.

Uber: Thanks for getting in touch, Kai. We’re sorry to hear about your trip experience. Please share any additional details. Feel free to add any questions or concerns you may have. We’ll reach out to let you know how we can help.

Me: Hi! I was imagining I would be offered a refund for the trip. I chose uberx instead of uberpool because I didn’t have time to go out of the way and make additional stops. The trip was estimated at about eighteen minutes and took about thirty in total, and I was late to my meeting as a result of the meandering route taken.

Uber: Thanks for letting us know, Kai. I am so sorry to hear that you got late for your meeting. We’ve reviewed this trip and adjusted your fare to $13.78. The difference in fare will be returned to your account in the next 1–3 business days. You’ll also receive an updated receipt by email. We’re sorry to hear this trip’s distance and time were longer than you expected.

Me: What? The uberpool would have been $4 and I paid $14 for uberX instead because I didn’t have time for extra stops!

Uber: When you select your vehicle option, the price displayed in the app at the beginning of the trip is the fare you’ll be charged for the ride. These fares are calculated dynamically, based on your pickup and drop off locations. Please note that your upfront fare price will include any higher rates that come into effect when demand is high. I can confirm that you were correctly charged the amount shown when the ride was requested, so you should be all set.

Me: Hi! My point is I didn’t get the service I expected. I paid $14 instead of $4 in order to get a ride with no stops. At most I should be charged the price it would have been for uberpool since the driver was making detours and stops on the way.

Uber: Hi Kai, We’re sorry to hear about the unexpected extra time on this trip. We know that when you request a ride with Uber, you trust that your driver will get you to your destination safely and comfortably. We have reviewed this trip. The fare you were charged is within our estimate for a trip from your pickup location to destination. As a result, the fare was not adjusted. That being said, we appreciate you taking the time to let us know there was an issue with this drop-off on this trip. Feedback like this help us improve the Uber trip experience.

Me: Hi! Would you agree that the primary difference between UberX and UberPool is that in UberPool there are extra stops along the way, while in an UberX the driver proceeds directly to your destination without detours or extra stops?

Uber: Thanks for getting in touch, Kai. We’re sorry to hear about your trip experience. Please share any additional details. Feel free to add any questions or concerns you may have. We’ll reach out to let you know how we can help.

Me: Hi! This is making me a tiny bit crazy. I already got previously in the thread the prewritten response you just sent. What I asked was a yes/no question. “Would you agree that the primary difference between UberX and UberPool is that in UberPool there are extra stops along the way, while in an UberX the driver proceeds directly to your destination without detours or extra stops?” Please answer with yes or no whether that would be the normal expectation.

Uber: We have reviewed this trip. The fare you were charged is within our estimate for a trip from your pickup location to destination. As a result, the fare was not adjusted. That being said, we appreciate you taking the time to let us know there was an issue with this drop-off on this trip. Feedback like this help us improve the Uber trip experience.

Me: Hi thanks, you’ve already sent me that one too. Would you agree that the primary difference between UberX and UberPool is that in UberPool there are extra stops along the way, while in an UberX the driver proceeds directly to your destination without detours or extra stops?

Uber: Thanks for getting in touch, Kai. We’re sorry to hear about the unexpected extra time on this trip. When you select uberX as your vehicle option, the price displayed in the app at the beginning of the trip is the fare you’ll be charged for the ride. We cannot take responsibility for external circumstances that may cause a ride to take longer than expected. Heavy traffic, road construction, and other factors that add unanticipated time to a trip are beyond our control. We appreciate your understanding. Please let us know if you have any questions.

Me: Hi there! Are you allowed to you send a response that isn’t pre written? I literally can’t believe I haven’t gotten a yes or no answer to a yes or no question despite asking it three times. PLEASE let me know, would you agree that the primary difference between UberX and UberPool is that in UberPool there are extra stops along the way, while in an UberX the driver proceeds directly to your destination without detours or extra stops? Also — is a driver making an extra stop an external circumstance or an internal circumstance? Three yes/no questions here.

[three days pass]

Me: Hi hi! Have you given up on me?

People say the challenge with automation relative to the human brain is its inability to handle unexpected situations. Handing unexpected situations is at least a large part of why customer support exists in the world, and studies show that when it comes to inquiries, all the meat is in the long tail. VCs, stay thy vengeful hand. And here’s to our species! – each of us individually the beautiful possessors of the most advanced neural nets known to mankind.

Very early on, our customers surprised us by using it as a budgeting and cash management tool also, and as a way to administer trusts or protect medical benefits, to put financial plans into effect, or sometimes as a way for the kids to be involved in their parents’ financial life without taking over. We learned a lot about their lives, perhaps more than we had expected.

One key thing we learned is that once you’re retired, budgeting and investing are tied closely together. Your savings aren’t an abstraction for the distant future — they’re how bills are paid. Over four years processing millions of financial transactions and being the actual form of payment for tens of thousands of customers, we learned a lot about how they spent the money, and where that money was coming from. To help them with budgeting, we needed to learn about their investing.

What we found was often deeply troubling.

We found that the financial advice retired folks were getting was often sales-driven or hunch-based rather than service-driven and objective. Most of our customers didn’t know how much they were paying in fees and what they were getting in return, how the investment option recommended for them correlated with their personal needs, or what other strategies had been analyzed and what tradeoffs were considered. Many of them were paying more than they were benefitting and not optimizing their income for the retired years of their lives.

Perhaps more importantly, the senior customer was treated almost as a leftover — “you put your retirement savings in a target date fund with us and guess what, now you’re past your target date” — as if they were a carton of cream going bad in the fridge. Today’s retirees are not past their target date, they are looking toward the future — and they absolutely know when they are being sold warmed-over retirement savings products rather than something built specifically for their needs.

Rather than watch retirees settle for mediocre investment options, we decided to do something about it — to help create a better future for the people who, incidentally, hold almost half of the wealth in America. We built True Link Financial Advisors and spent the past year with around two hundred clients in California, learning and refining our approach, bringing in the best of what’s available, throwing away the worst of it, and adding unique benefits specifically for people living off their assets. We built the kind of investment firm we’d want managing our own families’ retirement savings… and then, put our own families’ money in right beside yours.

Today, we’re making that investment approach available to the public for the first time.

The core of our investment approach starts with an understanding of a person and what they need. We start by looking at factors like age, assets, health, insurance coverage, and your current monthly budget and inbound cash flow. Based on this we figure out when you’ll actually need the money — and use those time horizons to recommend investments in low-cost, broad-based index funds that balance minimizing financial risk with providing returns that will meet your financial needs.

Our approach is algorithmic — not based on hunches or personal preferences — and developed by an investment team that has collectively managed hundreds of millions and even billions of dollars. We try to make it clear exactly what the tradeoffs are — more in stocks or more in bonds, earlier or later retirement, reducing or increasing spending in the near term, including annuities in the financial plan or not.

Unlike many other investment advisors, we’ll tell you for free exactly how we’d invest the assets — dollar amounts and ticker symbols — and you can print it out, take it home, and invest it yourself, never paying us a penny. We think if everyone did this it would make investing a lot easier, cheaper, more objective, and more transparent.

We also offer an advising service, where if you choose to hire us, you’ll get a human as well as an algorithm. We’ll tell you in plain English exactly what that human will do to earn the fee, and exactly what that fee will cost. And we encourage you to consider alternatives — look at what we say we’ll do, and consider hiring a different financial advisor, using an online service, doing it yourself, or asking a friend or family member to help you with it. Making it easy to comparison shop would improve both service and value across the financial industry.

Most of all, we hold ourselves to a fiduciary standard. Our financial advisors don’t earn commissions or have sales targets, we don’t get kickbacks or referral bonuses from third parties, we don’t charge confusing fees, and we don’t sell proprietary products.

To provide good service to retirees, a financial institution needs not only high ethical standards and sound financial systems, but also an understanding, empathy, and appreciation for the life situation of the customer. There’s a need for a financial institution that offers safety, support, and independence.

From our very first existence — as a living room wall covered in sticky notes — we’ve described our vision as “The financial services brand for older adults.” As we reflect on what that means, we mean both that we want post-retired adults to come to us and to continue choosing us as their trusted partner every day. But we also mean for in its metaphorical sense — we want to be for retirees, acting in their favor as advocates and entrepreneurs, making changes to an industry that often serves them poorly and almost always could do better.

We hope by making our financial toolkit available to the public for the first time retirees are able to make more objective, informed investing choices, get better advice and better service, with lower, clearer fees. We’re here to change the status quo — and bring better financial services to a population that has earned a better shake than what they’re getting.

Kai Stinchcombe is CEO of True Link Financial, Inc.

Why True Link is launching an investment division today was originally published in The Policy on Medium, where people are continuing the conversation by highlighting and responding to this story.