Is It Safe to Use a New Payment Service? How to Judge One With No Track Record

by PayFi · 5 October 2026 · Notes from building

Every payment service has a first day, and on that day it has no history to show you. Years in business feel like the best proof of safety, but they measure survival, not how a company treats your money. On day one, the better question is simple: what happens to my money if this service disappears tomorrow, and can I check every payment myself?

A service that launched this morning

Let's meet a seller we will call Bima. He is not a real person, but his hesitation is one every new tool has to face.

Bima sells short study guides for university statistics courses. On the morning of October 1, 2026, a friend sends him a link to a new payment service called PayFi. You paste your wallet address and a link to your product, it says, and you get a payment link. Buyers pay in USDT or USDC, digital dollars, and 99 percent goes straight to your wallet.

Bima likes the idea. He has buyers abroad who struggle to pay him. But then he notices something: the service launched that very day.

"Launched today," he thinks. "No reviews. No history. Nobody has used it. What if it has bugs? What if it disappears next week with my money?"

These are good questions. And they are the same questions a buyer asks when they meet an unfamiliar checkout page for the first time.

On the other side of Bima's problem is Laila, a student in Kuala Lumpur who wants one of his guides before an exam. Her bank card is often declined by foreign shops, but she has some USDT on an exchange. When Bima sends her a payment link from a service she has never heard of, she has her own doubts. Is this page real? If she pays, will the guide actually open? If something goes wrong, who will help her?

Neither of them can look up years of reviews. Both of them still need to decide.

This essay is about how to answer them when the usual answer, "it has been around for years", is not available. It uses PayFi's own first days as the example, because they happened in public and can be checked.

Why track record feels like the answer

When we decide whether to trust a business, age is one of the first things we look at. A shop that has been open for twenty years has probably treated its customers well enough to survive. A brand-new one has not proved anything yet.

That instinct is not wrong. New businesses really do fail more often than old ones. The US Bureau of Labor Statistics followed private business establishments that opened in March 2013. After one year, 79.6 percent were still operating. After five years, 50.6 percent. After ten years, 34.7 percent (BLS).

So roughly one in five new businesses does not reach its first birthday. If you are about to hand money to a service that started this morning, it is reasonable to wonder whether it will be there next year.

But notice what that number measures. It measures whether a business survives. It does not measure whether your money is safe while you use it. Those are different questions, and confusing them leads people astray in both directions.

Age is not the same as safety

Some of the most painful losses in recent years came from companies that were old, large and well known.

FTX had been running for years and was one of the biggest crypto exchanges in the world when it collapsed. It had told customers their deposits were kept safe and separate. According to the US Department of Justice, its founder was sentenced to 25 years in prison for "stealing over $8 billion of his customers' money" (U.S. Department of Justice).

Celsius, a crypto lending platform with years of history, told customers in May 2022 that it was "stronger than ever". The US Federal Trade Commission says it made claims like these days before freezing customer accounts, and that it misappropriated more than 4 billion dollars of consumer deposits (FTC).

Wirecard was a German payment company listed on the stock market and checked by a major audit firm for years. In June 2020, it admitted that 1.9 billion euros it claimed to hold most likely did not exist (Al Jazeera).

Each of these companies had a long track record. What they also had in common was custody: they held large amounts of other people's money, for long periods, out of sight. Their customers' safety depended on the company's honesty, and nobody outside could check it until it was too late.

So age is weak evidence in both directions. An old company can still lose your money. A new one does not have to be risky, if it is built so that it cannot lose much of it in the first place.

The better question for day one

If track record cannot answer the question, what can?

Here is a more useful question to ask about any payment service, new or old: if this service disappeared tomorrow, how much of my money would disappear with it?

For a service that keeps a balance for you, the answer is: whatever is in your balance. If sales collect for two weeks before a payout, that is two weeks of sales. PayPal, for example, says it may hold a new seller's first payments for up to 21 days (PayPal); that is three weeks of sales sitting with someone else.

For a service that sends each payment on to you right away, the answer is much smaller: only the payments in progress at that moment.

A little arithmetic makes the difference concrete. Imagine a seller who makes 20 sales a day at 10 dollars each. With a payout every 14 days, up to 2,800 dollars of their money can be sitting with the payment company at any time. With each payment split to their wallet within seconds, the amount in someone else's hands at any moment is usually one sale, perhaps two: 10 or 20 dollars. Same seller, same sales, a very different answer to "what if they disappear?"

The second question is just as practical: can I check each payment myself, without asking the service? If every payment leaves a public record that you can look up on your own, you do not need years of reviews to know whether the service did what it said. You can see it, payment by payment, from the very first one.

These two questions do not require any history at all. They work on day one.

What PayFi looked like on day one

PayFi opened to the public on October 1, 2026. Here is how it answered those two questions from the start.

How long it holds money. Every order gets its own payment address, made only for that order. When the buyer's payment arrives, PayFi checks it and, in a single Solana transaction, sends 99 percent to the seller's wallet and 1 percent to PayFi. That usually happens within about 20 seconds. There is no balance, no payout schedule and no account to collect money in. If PayFi vanished, money already split would already be in sellers' wallets.

What it does not hold. It does not ask for names, emails or IDs. There is no customer file to leak, and no account to freeze.

Whether payments can be checked. Every payment, and every split, is a transaction on the Solana blockchain. Anyone can look it up on a free blockchain explorer such as Solscan, which PayFi does not run.

There is also an honest detail about those first days. PayFi's public ledger, the open book of every payment it settles, did not exist on the morning of October 1. It was opened a few days later. Its first line records how many payments had been settled since launch, and says: "From here on, every one is written down." Every payment from launch day onward, including earlier test payments, was then written into the book with its original date, so nothing from day one is missing from the record.

Day three: a real payment, replayed

Promises about speed are easy to make. Here is one that can be checked.

On October 3, 2026, a buyer paid 10 USDT for a product called Basics of Statistics. The buyer used a swap service to send the money, and that service took about 30 minutes before the payment arrived; that part had nothing to do with PayFi. Once it arrived, the timing was short. The payment landed at the order's address at 07:35:30 UTC (Solscan). At 07:35:35 UTC, five seconds later, 9.9 USDT reached the seller's wallet and 0.1 USDT went to PayFi, in one transaction (Solscan).

You do not need to trust this paragraph. Both links open the real transactions on Solscan. You can see the amounts, the wallets and the times for yourself.

That is the point. A service with three days of history cannot show you three years of reviews. It can show you that every payment it handled did exactly what it said, and let you verify that without asking it.

Day four: an invoice, paid in fifteen seconds

The next day brought a different kind of payment.

On October 4, a client paid an invoice of 10.1 USDC for a logo conversion job, sending it from their own wallet. The payment arrived at 02:31:42 UTC (Solscan). At 02:31:57 UTC, fifteen seconds later, 9.999 USDC was in the freelancer's wallet and 0.101 USDC went to PayFi (Solscan). The client kept a receipt with the memo, the amount, the freelancer's wallet and the transaction code.

Again, the evidence is not a testimonial. It is two public transactions anyone can open.

Day five: the book sealed into Bitcoin

On October 5, the public ledger was sealed into Bitcoin. In plain words, a short fingerprint of the whole book up to that point was recorded in Bitcoin block 969,936, using a free tool called OpenTimestamps. From that moment, no line up to that point can be changed without the change being visible to anyone who checks.

For a service that was four days old, that is an unusual thing to have. It means that the first days of its history are fixed in a public record that nobody, including PayFi, can quietly rewrite. Whatever happens later, those days can always be checked.

Reading the first page of the book

It is worth seeing what a very young public ledger looks like, because it shows what "checkable from day one" means in practice.

On day five, PayFi's book is short. The first line, the genesis, quotes a newspaper headline only as a date marker, records a recent Solana block that could not have been guessed in advance, and states how many payments had been settled since launch. After it come the lines for each payment: the price, the amount that reached the seller, the fee and the Solana transaction. Lines from before the book was opened are marked as written in afterwards, with their original dates. Then come the seals: entries showing that the book's fingerprint was stamped into Bitcoin.

A summary at the top adds it all up: how much came in, how much went to sellers, how much was kept as fees, how much was refunded, how much is waiting for its owners, and a final line for anything unaccounted for, which should be zero.

For a stranger deciding whether to use the service, this first page is more useful than a hundred reviews. Reviews tell you how people felt. The book tells you what happened to every cent.

What a new service cannot promise

An honest look at day one also has to say what newness really means. A brand-new service has risks that age would reduce, and pretending otherwise would be the kind of "trust us" this essay argues against.

Bugs. New software has not met every situation yet. A strange payment, an unusual wallet or a busy network might reveal a problem nobody has seen. Clear refund rules help, but they do not make bugs impossible.

Downtime. If the service is offline, payments sent to an order's address wait there until it comes back. They do not go to a third place, but they are not split to the seller until the service is running again.

Money in progress. For the seconds or minutes between a buyer's payment arriving and the split, the money sits at the order's address, which PayFi's system controls. That window is short, but it is real. If PayFi disappeared permanently in the middle of an order, the money for that order could be stuck.

The small scale. PayFi is run by a small team on a small server. That keeps costs low, which is how the fee can stay at 1 percent, but it also means there is no large support team.

Things no payment service controls. The companies that issue USDT and USDC can freeze specific addresses. The value of a stablecoin depends on its issuer. These limits apply to any service that uses them, new or old.

Listing these is not a reason to avoid a new service. It is a way to choose how much to put through it while it is young, with open eyes.

How a seller can judge any new payment service

The questions that worked for PayFi's first days work for any new tool. Here is a short list a seller can use on day one of anything.

How long does it hold my money? The shorter, the better. Seconds is very different from weeks.

Where does each payment go, and can I see it? Look for payments you can check on a public record, not just on the company's own dashboard.

Are the fees and rules written down before I start? Look for a fixed fee and published refund rules, not "contact us".

What do they ask me for? A service that asks for little has little of yours to lose or leak.

What happens if they disappear? If the honest answer is "your balance goes with them", think about how big you let that balance get.

Can I test it small first? Coinbase advises making "a small test transaction" before sending larger amounts to a new address (Coinbase). The same idea works for a new payment tool: create a link for a cheap product, buy it yourself or ask a friend to, and watch where the money goes.

How to run your own first test

A test payment sounds technical, but it takes about ten minutes, and it answers more questions than any article can.

First, create a payment link for something cheap: a short guide, a sample file, even a thank-you note with a receipt. Use the wallet where you want to be paid.

Second, pay it yourself, or ask a friend to. If you have USDT or USDC on an exchange, withdraw the exact amount on the Solana network to the payment address the checkout page shows. Watch the page while you wait. It should notice the payment by itself, without you pasting anything.

Third, check your wallet. The seller's share, 99 percent of the price, should arrive within a short time after the payment lands.

Fourth, look it up. Open the transaction on Solscan and read it: the amount leaving the order's address, your share arriving in your wallet, the fee going to the service. Then, once the public ledger has caught up, find the same payment there.

If every step matches, you have learned something no review could tell you: that this service, today, does exactly what it says with your money. If anything does not match, you have learned that for the price of a cheap test.

How much to put through a new service

Even after a good test, it is sensible to let trust grow with use. But the way to do that depends on how the service handles money.

With a service that holds balances, caution means keeping the balance small: withdrawing often, not letting weeks of sales pile up. Your risk grows with every day you leave money there.

With a service that splits each payment straight to your wallet, there is no balance to manage. Your risk at any moment is only the payments that are in progress, usually for seconds. That changes how you can grow. You do not have to keep withdrawing to stay safe, because there is nothing to withdraw. You only need to keep watching that payments keep arriving the way they should, which the public record lets you do at any time.

For buyers, the same logic applies one purchase at a time. Start with something small from a seller you know. Check that the product opened and that the payment shows up where it should. Then decide how you feel about the next purchase.

How a buyer can judge an unfamiliar checkout

Buyers meeting a new checkout page can use a shorter version.

Is the price clear, and is it the price you agreed? If the page shows a different amount from the one the seller told you, stop.

Does the payment go to an address made for your order? A page that gives you one address per order, and tells you that the address is for this order only, is behaving like a payment system. A random wallet address pasted in a chat is not.

Do you get something to keep before you pay? A private order link, a receipt or a reference you can use later is a good sign.

Are refunds explained before you pay? Look for written rules for paying too much, too little or too late, not just a support email.

Can you check your payment afterwards? A transaction code you can look up on a public explorer turns "I hope it worked" into "I can see it worked".

Laila's side of the test

Bima's friend in the test was not imaginary for long. He sends the link to Laila, the student in Kuala Lumpur, and tells her honestly that the service is new.

Laila opens the link on her phone. The page shows the guide's name and the price, 10 USDT. It asks her to tick a box confirming she will send on the Solana network, and then gives her a private order link to save before paying. Only after that does it show a payment address, with a note that it belongs to her order only.

She withdraws 10 USDT from her exchange to that address. Her exchange takes a few minutes to send it. When it arrives, the page changes by itself: payment received, product unlocked. She downloads the guide.

Afterwards, out of curiosity, she copies the transaction code from her exchange into Solscan. She sees her 10 USDT arrive at the order's address, and by clicking the order's address, a second transaction sending 9.9 USDT to a wallet whose address matches the one Bima shows on his page.

She did not need to trust a brand she had never heard of. She needed a page that behaved predictably and a payment she could see for herself.

Why build in the open from day one

There is a reason PayFi's first days are visible in such detail. It follows from how the service is built.

A service that asks for no account, no name and no ID cannot also ask for blind trust. It has no famous brand and no years of history to lean on. The only fair way for it to earn trust is to make itself checkable from the very first payment: hold money for seconds rather than weeks, put every payment on a public record, write the rules down, open the books and seal them where they cannot be changed.

Building that way from day one has a cost. Mistakes are visible. Early numbers are small and public. There is nowhere to hide a quiet first week. But it also means that the first days do not need to be taken on faith. They are there for anyone to read.

What time will add

None of this means time does not matter. A service that has run for years has met more strange payments, more network problems and more mistakes, and has had the chance to fix them. That experience is real, and a new service does not have it yet.

But for a service with an open book, time adds something more specific than reputation. Every day adds more lines to a public record that cannot be rewritten. A year from now, PayFi's track record will not be a collection of reviews or a claim on a website. It will be the book itself: every payment since day one, each linked to its transaction, each sealed into Bitcoin.

That is a different kind of track record. It cannot be bought, edited or polished. It can only be added to, one honest line at a time. And anyone can read it from the first line, which is exactly why the first line matters.

Your own day one

There is a quieter lesson here for sellers, because every seller also has a day one.

The first time you sell something online, you have no reviews either. Buyers who find you have no history to judge you by. You face the same question PayFi faced: why should anyone trust something new?

The answer can be the same. Instead of asking buyers to trust you, give them things they can check. Show your work before they buy. Describe exactly what they will get. Use a way of getting paid that gives them a receipt and a public record of the payment. Publish your wallet address in places that clearly belong to you, so they can see the money goes to you and nowhere else.

A track record is something you build over time. Being checkable is something you can offer from your first sale.

Questions people ask

Should I wait until a new payment service is a year old? You can, and some people will. But age alone does not tell you how a service treats money. A better approach is to look at how long it holds your money and whether you can check each payment, then start small.

What if a new service shuts down? With a service that splits each payment to your wallet within seconds, past payments are already yours. What is at risk is any payment in progress at the moment it stops. With a service that holds balances, your whole balance is at risk.

Is it risky to be one of the first users? Somewhat, in the ways listed above: bugs, downtime and a small team. It is also when you can see everything most clearly, because every payment so far fits on one page of the public ledger.

How can I check that the early payments were real? Open the transactions linked in this essay on Solscan, or open the public ledger and follow any line to its Solana transaction.

Does a public ledger make a new service safe? It makes it checkable, which is different. It means that if something goes wrong, you and everyone else can see it. That is not a guarantee, but it is far more than "trust us".

Back to Bima

Let's go back to Bima on the morning of October 1.

He decides not to wait a year, and not to trust blindly either. He creates one payment link for his cheapest study guide and asks a friend abroad to buy it. The money lands in his wallet seconds after his friend's payment arrives. He opens the transaction on Solscan and sees the split: 99 percent to him, 1 percent to the service, nothing anywhere else.

A few days later, he looks at the public ledger and finds his sale there, sealed into Bitcoin with everything else from the first days.

The old belief says that a new service is unsafe until it has years behind it. Sometimes that caution is wise. But the more precise belief is this: what makes a payment service safe is not how long it has existed, but how little of your money it holds and how much of what it does you can check. A service built that way can earn trust on its first day, one payment at a time.

Laila passes her exam. When her classmates ask how she paid a seller in another country with a card that never works abroad, she sends them Bima's link, and tells them the one thing she learned: you do not have to believe it works. You can watch it work.

Sources

• U.S. Bureau of Labor Statistics, "34.7 percent of business establishments born in 2013 were still operating in 2023": bls.gov
• U.S. Department of Justice, "Samuel Bankman-Fried Sentenced to 25 Years in Prison": justice.gov
• Federal Trade Commission, Celsius Network settlement: ftc.gov
• Al Jazeera, "What $2.1bn? Wirecard says missing funds most likely do not exist": aljazeera.com
• PayPal Help Center, "New PayPal account – payments on hold": paypal.com
• Coinbase Help, "Sent to wrong address": help.coinbase.com
• Solscan, October 3, 2026 payment: buyer's payment, seller's payout
• Solscan, October 4, 2026 invoice payment: client's payment, freelancer's payout
• PayFi Ledger: payfi.cx/ledger