Stripe Terminal – Programmable point of sale | Hacker News Reader

Stripe Terminal – Programmable point of sale

stripe.com

539 pointsanacleto217 comments

It's interesting to see:

- Square become Stripe, while Stripe becomes Square

- Stripe owning the whole user experience in purchasing

- If they own the physical end points, will they eventually cut out credit card companies? You can already see that they are pushing for collecting emails with checkout.js. The first step is to have direct contact with the end user.

- If they cut out credit card companies, what's to stop them from moving into personal payments (e.g. Square Cash, Venmo) and then banking?

Nothing stopping them. And I wish they would. The MC/Visa duopoly needs to be disrupted.

The MC/Visa duopoly needs to be disrupted.

Yes, but not by anything even vaguely resembling credit cards as we know them today. The whole system is insecure, inefficient and unreliable by design, and the credit aspect introduces an extra level of legal and financial mess, particularly for those people who just want a convenient payment method but don't actually need to defer settlement. Card payments may be the perfect example of the dangers of monopolies and oligopolies.

What, and replaced with another monopoly?

Or a duopoly of Stripe and Applepay?

>nothing is stopping them.

I'm guessing their Network agreements with Visa and MasterCard would stop them. You could issue a stripe card that only works with stripe merchants. But I can't imagine visa allowing their imprint on a card that can both process anywhere AND be used to slowly cut visa out of the picture.

Many European cards have both the VISA logo and also a national chip&pin debit scheme that has nothing to do with VISA.

Maybe you should take a look at the hard numbers before you talk about cutting out the CC companies.

Paypal is leading the game and they are not able to cut out CC companies.

I don't see much Paypal deployment in the physical world. The only place I've ever seen the option to pay with Paypal is at Home Depot. (I didn't try it because it sounded like you had to log in with your password and second factor... which did not sound like a good use of time in the checkout line.)

Or annoyingly, donations t9 websites like Wikipedia require Paypal

Maybe it depends on the country you're in, but you can definitely use methods other than PayPal. When I donate to projects with foundations in Europe, I usually just wire the money directly to their bank account. (I'm in Germany.)

Phone number and 4 digit pin, super easy actually.

I’m confused. I pay with PayPal balance all the time.

Good for you. Question: Do credit cards still exists, outside of your bubble?

You can't cut out the cc companies. Both Stripe and Square are riding on their rails. The best they can do is try to negotiate the rates down and either keep a higher percentage for themselves or pass on the savings to their customers.

I think the interesting idea is what happens when you use a Stripe issued card[0] on a Stripe API powered terminal. The system, somewhere, will know it is a Stripe card based off the number, but I'm willing to bet the CC networks are protected from being cut out. At least for now.

Great job team! Looking forward to more of this in the future \o/

[0] https://stripe.com/issuing

They're protected from being cut out because they are the networks on which the credit card system runs. If you look at a stripe-issued card, you'll see there's still a visa or mc logo in the corner.

I understand how the system works; the requests hit the API before the credit card network. Hence the ability to cut out the network if Stripe is willing to take the chargeback/fraud risk.

No, they would need to become money transmitters according to what you wrote above. AFAIK they aren't money transmitters yet.

It depends on the jurisdiction, but they are in many! It is something they have been working towards for quite some time.

They list themselves as having money transmitter licenses across the US.

see e.g. https://stripe.com/us/payments and https://stripe.com/connect (under "Rigorous Compliance" and "Compliance respectively").

You'll also find them on CA's directory of money transmitters. http://www.dbo.ca.gov/Licensees/money_transmitters/money_tra...

You're saying Stripe would process directly when their own card is used on their own network, and then outside the network they'd process through visa/mc?

Well something like that depending on the Stripe infrastructure.

That currently happens when an issuer card is used in a merchant that is acquired/processed with the same issuing bank. It is called an "on us" transaction and doesn't need to fire through the card scheme rails.

I assume that still incurs a visa fee. Perhaps just bypasses the underwriting. Otherwise, I can't imagine visa would allow this.

I can't see how they could disallow it to be honest. That would be seriously anti-competitive.

There's a whole infrastructure of liability for payer and merchant that exists below the surface of every transaction.

This is why, unfortunately, credit card companies can charge the fees that they do. For example, the issuing banks take on liability if payers don't pay, for example. I don't think Stripe is willing to get in the business of this unless they want to turn into a bank.

I think you might be surprised how much risk Stripe is willing to take on when they can gain hundreds of basis points on cutting out the networks. Only time will tell though :)

Not if it means having to turn into a bank. The repercussions on their business will be draconian. Although I would be surprised if they haven't already explored or fully gone through a money transmitter license for a subsidiary.

Can't they just buy a Bank?

Almost certainly they would, and there might be some small banks out there they could buy or merge with. That would get them out of a lot of legal issues, and also buy them some relevant legal expertise.

However, I seem to recall WalMart investigating this option a while back, and it never happened. It may be that the more you look at the option of becoming a bank, the more it looks like it would take over your life/business. Also, there are doubtless many lobbyists who the existing banks employ who would try to stop it.

I think the issue is being a bank vs owning a bank.

What's the risk if payments were to be debited from the customer's bank directly?

Related side note: I've always wondered, if people don't ever need the "credit" in the credit card transaction, could they get a discount of 2 odd % on each transaction that the CC companies charge, say if they connect their bank accounts directly to they payment methods? This would effectively a debit card but not using the VISA/MC networks. What's stopping Apple/Google pay from processing payments directly from my bank?

Almost none. PIN/passwords render fraud to negligible amounts. That's why debit card transaction have fees for ~$0.10 under Durbin. Companies like Stripe make a killing whenever their customers use debit cards instead of credit cards because they're charging 2.9 +0.30 with a cost of 0.10.

But if you had to rely on debit cards only, then commerce at least in the US would grind to a halt. I believe the average American CC debt is ~$8000.

nordstrom used to do this, back when it was also a bank and issued their own line of credit cards. they also negotiated more favorable terms with visa because of it. they marketed the crap out of their own cards so they could keep more of the fees.

but the regulatory environment tightened up and being both a bank and a retailer became less appealing, and consumers found store branded credit cards less appealing, so they sold the bank portion.

Yeah, you can. I'd expect PSD2 legislation will all-but destroy CC companies across the EU within 3 years (just for online use initially).

Strong Customer Authentication (SCA) requirements coming into force in Sept 2019 are making card payments increasingly unattractive.

Payment Initiation (via PISPs, which Stripe undoubtedly will become) will offer free, instant, seamless push-based payments where fraud is essentially impossible (and the merchant isn't liable anyway).

Imagine having Cuvva's payment button pop you out to the Monzo app, hit accept, then immediately pop back and have paid. No card networks or fees involved. Just a simple Faster Payment directly from the customer's bank account to ours.

For in-person payments, I'd imagine a similar arrangement could be facilitated via Apple Pay.

May take non-EU countries a while to follow, but they surely will eventually. The days of CC companies are numbered.

Your example of the payment button is exactly how Dutch-only iDeal works and has adapted well to mobile. The pop-out however is to your own bank's app/website, which works cause there are just 10 banks. Transaction fees? < 0.50 € per transaction

Credit cards will never disappear as long as they offer the ability to chargeback if the goods are unsatisfactory, and competing systems do not offer that.

In the UK, debit cards can chargeback. While this is currently implemented over MasterCard/Visa network, it's the bank that performs it and there's no reason to expect that would change if something replaced the plastic.

Something similar (but not quite identical) exists for direct debits called the direct debit guarantee.

They already are disappearing in the UK. None of the new banks are offering credit cards.

People have been saying that for decades and the CC companies are as strong as ever. It's proven to be a great solution and remains the best user experience.

Right but this is actually being caused by very recent legislative changes. Things have started changing very quickly - particularly in the UK.

I see all of your points, and agree things are going that way.

But, for me, I will be staying with the cards until the chargeback/dispute schemes match up on bank transfers. I don't want to have to go to the small claims court if I have a problem.

> The days of CC companies are numbered.

Some sibling comments echo that this is underway in Europe. In America, the credit part of "credit card" is relevant. Many purchases on CC's here are intended as short-term (or longer) loans. (Something like 40% of American credit card holders carry a balance.) Direct debit from a bank account does not (and cannot) fulfill this use case.

Not to mention the fact that people like me who get 2% cash back or points worth as much as 5-10% are going to want to keep using their same cards.

This went away in the EU already. Interchange is capped at 0.3% (0.2% for debit), so there’s no way to provide rewards for cardholders.

Sure it’s sad that we no longer get air miles etc. But the reality is that we never should have in the first place - was a symptom of a nonsensical system.

> I'd expect PSD2 legislation will all-but destroy CC companies across the EU within 3 years

I wish. PSD2 will do nothing for consumers and won't hurt CC companies in the slightest. The requirements to actually be able to use the PSD2 APIs are insane. The only companies who will be able to use them are big players like banks, CC companies, large payment processors and large IT companies like Google/MS/IBM/Apple.

They're still one step away from Square. This is Stripe Terminal and not Stripe POS. This competes more with Square Reader SDK. If I remember correctly Square Reader SDK only launched fairly recently.

Your point is still valid, but I just wanted to point out that building a POS like Square is a different beast.

I'm sure they've thought about it. However, as others have found, getting through the amount of fraud out there with a margin less than the ~3% that credit card companies charge, is really hard. I mean if you were perfect that would be 3% profit. If your uncaught fraud rate is even a few percent higher than that of companies who have been doing this for a long time, you will be bleeding money.

Again, I'm sure they've thought of it, but the credit card companies are actually pretty good at a very hard task, which is enabling super-easy and quick payment verification without getting eaten alive by fraud.

> without getting eaten alive by fraud.

Because fraudulent use of cards is charged back to the merchant, the card companies are less exposed to it than you might imagine.

Where they are exposed is if a merchant goes bankrupt without delivering, and the purchase was on credit card. The merchant account holder would then be liable

But, they have to maintain a system to deal with that. Also, if they pass all fraud back to the merchants, at the levels that would happen if they had no good fraud detection, then merchants would stop allowing credit cards. We take for granted that most transactions are valid. If the credit card companies did not invest a lot in automatic fraud detection, the percentage of fraudulent transactions would be something like the percentage of email which is spam.

They still leave much of the risk to the merchant. Turn on 3d secure, watch how the banks often don't even respond, and watch the conversions plummet as your hard won customers get caught in a whirling loop. Turn on all the fraud tools and watch how nobody ever checks-out again...

Ya, depending on your type of business, you really may not need any fraud tools at all. I deal in b2b SaaS. We had to work to get any and all fraud tools turned off because we have essentially zero fraud (companies that you work closely with tend not to try to pay their monthly recurring fees with stolen credit cards.)

This is a serious issue, especially for small merchants dealing with digital products. A chargeback is too much work to fight and rarely in favor of the merchant.

The average card present transaction charge in Australia for scheme cards is less than 1.5%, even for small turnover merchants. With higher turnover businesses getting to about 1%.

Banks don’t need to charge 3% to deal with fraud. Merchant agreements effectively make it their responsibility to bear the brunt of any chargebacks anyway.

For sure the banks make good money, but the question is whether or not a company that is new at it, would be able to make a profit on 3%. I submit that it is a difficult task, and a new company would struggle. Not saying it's impossible, just that it's hard.

It's more that U.S. companies have various rewards programs (travel points to cash back). One of my cards gets 2% back on nearly everything, and 5% back on their quarterly promotions. I tend to use it for most of may day to day spending, and try to pay it off every month. That's why the merchant fees are so high... much less an extent to fraud handling.

The card brands are nothing more but a network. That's it. They have no skin in the game. All fraudulent charges move like this

card issuer(bank) ---> processor (square) ------> merchant

Not quite true. First, some fraud gets eaten by the processor. Second, the merchant underwriter is on the hook for a merchant who goes bust before fulfilling what has been paid for.

Actually, if you want me to make it detailed it goes like this.

  card issuer(bank) ---> processor (square) ------>  ISO ------> merchant ---> customer
  |                              |                   |
  V                              V                   V
  card network                      underwriting bank

the only way the processor will ever eat the cost is if the ISO couldn't cover the loss, bankrupt. The only way the ISO can't collect from the merchant is the same or if the merchant stops processing. If the merchant can't beat the chargeback and isn't fraudulent, their only option is to sue the customer for fraud. But usually the ISO will take the merchant to collections. Disputes are like hot potatoes, everyone keeps tossing it till there's no one to catch it.

With that said the underwriting bank doesn't eat any of the chargeback on behalf of the merchant. The underwriting bank will only eat loss that they underwrote for a processor or an ISO acting like a PayFac

At least this much I know from my side of the industry, perhaps you work for a bank that eats the cost. I do like to know their name and become a customer!

Do you have a source for any of that? I don't believe ISOs take on any risk. They are "independent sales organizations" and they are a dime a dozen.

They are signing up companies directly to the merchant banks each with their own underwritten merchant account. And they are a dime a dozen. They aren't taking on any risk. Just the effort to onboard new customers.

I believe I mentioned ISOs that are Payfacs. Some ISOs want to act like processors, if they do, then they assume the risk. Some ISOs have their own underwritten, most banks are too conservative and will lose your business by being too cautious, you can underwrite yourself if you are big enough. I work in this domain ;) and have interacted with the plumbing from onboarding all the way to account closure.

You've confused me as well. An ISO is sort of definitionally not the merchant underwriter ("Independent Sales Organization"). Maybe you're saying some underwriters/processors who sell direct? What is an example of an ISO that is a PayFac? There aren't many players in the business so naming names might help the general understanding.

unfortunately, I'm not at liberty to name names. But see this https://www.pymnts.com/digital-payments/2017/wepay-cofounder...

I've always thought fraud would be much more preventable if credit card companies would simply send you a text or verification code each time your card was charged, or require a PIN.

It seems this is difficult for the big credit card companies and banks (which is why it hasn't happened), but I wonder if it might be easier for Stripe to incorporate something like this into their Issuing product.

I think it is/was difficult when the product was just a plastic card. I agree that, as the "card" is increasingly becoming a smartphone, something better should be possible.

My bank, Westpac, calls me up every time I make an online payment with my VISA card or when my card is used at point of sale or at an ATM in a country I haven't notified them that I'm visiting.

My bank will let through thousands of dollars in ATM withdrawals in a country I've never been too. Then lock my card up when I'm trying to buy $12 of groceries in my hometown and force me to call and wait on hold to get it unlocked.

similar here - from US, had a few trips to russia and china over a few years. never told the bank. went over there, used my cards, no issues, ever.

came back home. no issues.

bought $12 from a merchant in sweden - a merchant I'd purchased the exact service from multiple times before - entire account locked for 'fraudulent activity'. Checks bounced, payments blocked, etc. Annoying as shit. Took a couple days to sort out (happened on a friday afternoon, of course). They refunded all charges, but... damn. I think things are better now (this was... 2014 IIRC).

This is already a thing, for merchants that want it. It's called https://en.wikipedia.org/wiki/3-D_Secure

By bank implements it as approving from my phone's mobile app via push notification. It's really slick. Here's what happened when I bought something from PSN https://imgur.com/a/iZgY74b

It exists. Conversion rates on sites that use 3D-Secure in the US went through the floor, which is why merchants focusing on that market have largely abandoned it.

Seems this was downvoted a lot but in Europe it's practically the expected behaviour.

Modern banks (N26, Revolut, Monzo, etc.) will send you a notification through their app and then you use the fingerprint sensor on your phone or a passcode to approve the transaction. With ATMs they will check your location and block any ATM transaction that is not near to you, or in a different country. The limits are very customisable and you can do it on the fly.

Most of the time you just have to open your bank app on your phone and hit confirm. You can lock the card when you're not using it and unlock it when you do, so no need to wait for a replacement or call support; meaning that if you lost your card you can go into the app and lock it straight away, just in case you find it again or it's returned.

It integrates with 3D Secure in a lot of places so you can still press your finger on your phone to approve.

It's a much more pro-active approach to bank fraud than what you get when you can't let go of the legacy system.

I am continuously shocked by the amount of defense that people have towards Stripe. I remember mentioning Stripe's 3% fee being too high for me the other day at a community event and I was attacked by people defending Stripe and their fee. These are people that don't work there, and have no true involvement with them (other than using their APIs).

Sorry everyone, but you really don't understand that Stripe's position is literally the middleman of everything. Their "product" is the API. Simplifying and unifying what the various processors have done and marketing the hell out of it.

That 3% fee that everyone defends is truly almost pure profit for them. I currently work with another payment processor (one of the top three largest in the world) and our credit card fees are measured in basis points. For those that don't know... a basis point is one-hundredth of a percent. The fees we are charged by our processor isn't even whole percentage points, it is measured in basis points. Granted it is more complicated because there are different fees for AMEX, Mastercard, Visa, etc. But it is still all totaled out to fractions of a percent.

Keep in mind that my credit processor is still making money while charging me in basis points (and they have way more employees than Stripe does). Stripe is a competitor to this processor. So if this processor makes money charging fees in basis points, than Stripe could make money charging in basis points too.

And before people think that I work for Walmart. I'll just say we charge about 2M a month in credit cards which is relatively small.

I promise that Stripe is doing very well off their 3% charge. I am not saying that they aren't allowed to make good money, but I really hate when people defend them like they are barely scraping by. Stripe is laughing their way to the bank with your 3% processing fee. They laugh even harder when they read you defending their outrageous fees.

Uh, that 3% includes the amount they are passing on to the credit card company. Which, for a little guy like me, is around 3%. It is literally no more expensive to go with Stripe than to roll my own.

Yes. Years back we moved from Elavon to Stripe since 2.9% on each transaction sure beat uncertain 1.9-5% we were getting based on what card customer was using (corporate and point cards carried bigger fee).

Stripe might be taking 30% of that 3% and passes 70% to payment networks. please look at adeyn balance sheet(https://www.adyen.com/investor-relations)

Well I would assume Stripe gets a better deal than a little business would, and I would expect them to take some significant cut, otherwise they would not be a viable long-term provider, yes? I don't want to be exploited mercilessly, but I also don't want to rely on a payment processor that makes no profit, it seems like they would not stay around for the long term.

Exactly. I signed up with stripe maybe 5 years ago. At that time, I was trying to find a processor, and came close to signing with a bank. The bank wanted me to sign a long-term agreement, the agreement had disadvantageous and confusing terms with unclear costs, and it was more expensive than stripe at baseline. The sales process was really sleezy: I only discovered all those terms because I read the contract carefully before signing.

Before that I used two other processors, both of which had a confusing mix of monthly and per-transaction fees.

If you are a very, very large business, maybe you can do better than stripe. But if you are small or even medium size, I am skeptical there is a better deal out there.

Even a small business can easily do better than A (when it comes to fees). The only time stripe would be a good deal is if you are processing almost exclusively high end corporate cards like Amex or corporate purchasing cards.

Stripe and others like them is also the best rate if you are processing very small annual amounts. Because every merchant account has around $10 in monthly fees, minimum. If you are not paying a monthly fee, either your processor is paying it, assuming that they will make enough on the spread to make up for it, or they are running your processing under their own merchant account and taking on the risk. Which is likely what stripe does until an account is large enough to warrant moving it under its own merchant account.

Here is an example of a relatively expensive rate for processing with a merchant account (2% plus 10 cents per transaction, with a 10 monthly fee) versus Stripe. Note how the breakeven for getting a merchant account is when you start processing more than $428 a month(assuming an average transaction is $15).

https://www.desmos.com/calculator/zuzn1lhfdg

My experience in the U.S., admittedly as a very small size merchant, is that 2% +10 cents is considerably better than I can get with a bank.

Banks have 2nd tier offerings for credit and debit processing that don't focus on gaining customers with low rates. There are a handful of platforms, First Data, Tsys, Elavon, Worldpay/Mercury & Heartland, and everyone who sells merchant services is selling you access to one of those platforms.

I was casually chatting with a small business owner about five years ago and he was raving about how much money moving to Stripe saved him, especially on AMEX cards.

stripe gets probably 0.8% + 5¢

I would love to know where you are obtaining an average sub 1% processing cost for US cards. USA has some of the highest interchange fees in the world.

I wonder how this come about. Interchange in Australia is way lower, you can get rates around 1% without too much effort and interchange plus rates even lower on average.

You would think a larger market like the US would be lower not higher because of the volume.

Is chip and pin ubiquitous in Australia? That would lower fraud rates. Also, the EU has a 0.3% cap on interchange. Average in US is 2%. My guess is Australia has a cap as well. So the answer is, lobbying.

Yeah turns out we do have a cap:

The weighted-average credit benchmark of 0.50 per cent will be maintained.

The weighted-average interchange fee benchmark for debit cards will be reduced to 8 cents per transaction, which will apply jointly to debit and prepaid cards in each scheme.

Interchange fee caps will be supplemented by ceilings on individual interchange rates: 0.80 per cent for credit; and 15 cents, or 0.20 per cent if the interchange fee is specified in percentage terms, for debit and prepaid.

To prevent interchange fees drifting upwards in the manner that they have previously, compliance with the benchmark will be observed quarterly rather than every three years. A scheme will be required to reset its interchange schedule in the event that its average interchange fee over the previous four-quarter period exceeds the benchmark.

Also chip and pin is ubiquitous.

Hey if Stripe really is reaping in the profits I have great news for you! Someone, maybe even you, can launch a competing project, offer a similarly polished API, and charge 2.5% and reap slightly less profit but still almost pure profit! Once that happens, Stripe will be forced to reduce its fee and everyone on the internet will be able to keep a little bit more of their money. Everyone wins (except Stripe and NewCo)! In the meantime, people get to choose if paying Stripe 3% is worth all the headache that it saves or if they are better off with a different approach. Hopefully not to many people will get tricked into pay Stripe outrageous fees for being a middleman while we wait for capitalism to work.

This is wrong. Stripe passes 2% or more of it's 2.9% on to Visa/MC and the card issuers. And larger Stripe merchants can negotiate that down as well as enter into an "interchange plus" arrangement which would be measured in BPs.

>>> I promise that Stripe is doing very well off their 3% charge.

I will tell you a secret. Stripe is not doing very well because they have a small volume of payments, all clients combined. A percentage of peanuts is peanuts.

I looked at them when I was working at a startup 2 years ago. The volume of transactions we were processing yearly was a 2 digits percentage of what Stripe was processing yearly across all their clients.

Needless to say, they're not adequate as a payment processor for any medium or large company.

The numbers quoted here[0] are soft and from a year ago, but at that time, Stripe claimed to handled handles tens of billions of dollars in internet transactions annually. If you were doing 10+% of that, kudos to you!

[0 ]https://www.bloomberg.com/news/features/2017-08-01/how-two-b...

Thank you. We were doing rather well by some metrics. Enough for the CEO to pay his first jet and for the yearly profit to fill a swimming pool in $1 bill. Not that we did that last part.

Humor aside. A few billions is actually fairly little for many businesses, remember that a typical margin is only a few percents of revenues (customer payments). A billion dollar company with a 1% margin can only afford 100 employees. You wouldn't be impressed if I told you that we were a 100 people company, would you?

Competing payment providers and banks are dealing in orders of magnitude above Stripe.

I would be impressed if a company had 100 employees today and planned for exponential headcount growth (say, 50% growth per annum). And had taken $440 million in funding. And had a product that wiped the floor with the competition.*

* YMMV, but for this dev, Stripe was a godsend.

Ya, for any dev, stripe is amazing. But for a CFO it's a money pit.

Stripe is great for small websites in the US.

It's disappointing for global payments and for high volume. I wish all developers here to outgrow Stripe.

Hrm, we accept payments globally in over 135 currencies and have special enterprise plans, like high volume pricing—but I'd love to hear more on what we should be doing better. Feel free to email me at edwin@stripe.com

Yes, Stripe's product is literally playing the middleman, but there are a lot of contexts where a middleman is nevertheless a net positive. That said, are your numbers for just the network fees or do they include interchange fees as well?

In any case, I remember the sort of hoops I had to jump through back around 2007 to setup a gateway through Authorize.net and a merchant account for a very small service (annual transactions around 150-200k). Every step of the process, from setting up the accounts to dealing with their API was a nightmare, and recurring billing threw some additional hurdles in for fun if I'm remembering correctly. Beyond that, there are a lot of ways to screw up with billing and payment processing. In my experience, Stripe's generally competitive and easy to deal with.

That's not to say that they make sense for all scenarios. They don't. But if you're a startup looking to make it to launch, there's a strong argument for keeping the billing side of things as straightforward as possible.

1) Stripe pays around 1% give or take 25 basis points depending on card type and location.

2) I highly doubt you’re getting sub 1% rate in the US with only $2M monthly, especially if you do card not present transactions. (See sources, I won’t call you a liar, because you could be in a very tiny niche or not US based, but all signs point otherwise).

3) most folks who use Stripe aren’t doing $2M in volume and likely don’t have the business credit rating / history to get a very low take rate. Further, Stripe doesn’t nickel and dime customers with other fees, so for lower volume customers the 3% actually might be a savings. If not, it’s a low markup for ease of use.

4) was the hostile attitude really necessary?

Sources:

https://usa.visa.com/dam/VCOM/global/support-legal/documents...

https://usa.visa.com/dam/VCOM/download/merchants/interlink-i...

https://www.mastercard.us/content/dam/mccom/en-us/documents/...

Are you in the US? Because interchange fees (which are the same whether you are Walmart or a hair salon) average about 2%.

Unless stripe is processing mostly generic debit cards, then they are most certainly not pocketing that full 3%.

I would guess that the makeup of stripe customers is similar to the US average, so they are probably paying close to 2% and then pocketing the 0.9% plus most of the 30 cents per transaction.

We are in a country where Stripe isn't available and pay 5.5% to our local processor here (and no fixed fee). So yes 3% is pretty good. They also offer custom pricing depending on your business model (e.g. micropayments).

Interesting point. Out of curiousity, is it ok to tell us what country that is? I am always interested to hear about how these things work differently in different parts of the world.

Sure, we are in Costa Rica.

> If they own the physical end points, will they eventually cut out credit card companies?

I seem to recall a story on HN around 6 months ago about Mastercard (might have been Visa) becoming an investor in Stripe, so I they'll do whatever they need to to keep Stripe in check.

It was Visa: http://fortune.com/2015/07/28/stripe-visa/

I can't help but think Square ceded the online space to Stripe despite that probably 50% or more of Square's merchants process online with another vendor.

I believe Stripe's original plan when they applied to YC was to start a bank.

We spent a lot of time building a terminal at Sidestep on top of Stripe (and a few other processors). Everything on the in-person payment level was done through an iOS or Android device. The hardest part wasn't managing devices (they could log in on almost any device), but the hardware integration for accepting credit cards. It exists in various places, but none of it could do exactly what we wanted to do. CardFlight was the closest to being a stripe friendly api for swiping cards. Usually the devices for swiping cards were too expensive and the failure rate was high. I don't think swiping hardware has gone down, unless you're using Square.

But hardware costs aside, the ability to fully integrate with the API on a terminal level is the most powerful thing here. It allows for so many cool POS integrations like the one we building at Sidestep.

To me, the coolest thing about this is that it supports Stripe Connect so you can create a POS that's used by others without having to deal with any money transfer yourself. Customer swipes card -> your software -> client's bank account.

how much fee is collected like this? Stripe Connect takes a fee, Stripe Terminal takes a fee, etc.

I don't think the fees stack like that. The terminal page says 2.7% + $0.05 per transaction. I believe that Stripe Connect doesn't take an extra fee, at least not in my experience.

It does. It takes a percentage of the connected user’s sales AND a $2.00 monthly fee per user AND the normal transaction fee.

If there was a competitor to Connect, I would use it. Fees end up being around 3.4%. That’s pretty hard to sell customers on.

So 3.4% for card present? That will not fly except with the smallest of businesses. (which is not where the profit is).

wha are some "cool POS integrations" ? what use cases do you for see making a big splash with this?

(I work at Stripe)

We are excited about the use cases of Connect platforms integrating Stripe Terminal! We have seen a lot of platforms like AtVenue and Mindbody building an end-to-end solution for their users -- completely on Stripe.

A shot across the bow against Square. This will be very interesting. Payments has always been a race to the bottom, I wonder how quickly rates will drop for POS payment providers now that Stripe is going after that business and directly against Square.

I'm surprised either Square or Stripe would do well in the "card present" space. 2.9% + 30 cents is pretty big markup if you're "card present".

Edit: See comment below, apparently the 30 cents is now 5 cents for card present. Helpful, though doesn't necessarily close the whole gap, depending on transaction mix.

Could you clarify? Is this markup only on card present transactions?

It's specific to the US. Within the US, if have your own merchant account, "card present" is a lot cheaper than "card not present".

So far, companies like Paypal, Square, and now Stripe, don't pass those savings on. They all have one rate that doesn't differentiate.

More importantly, fraud rates are much lower with CP than CNP. If the Visa/MC in the US goes with a PIN-type system like the rest of the world, fraud rates drop to near zero, which makes Card Present insanely profitable for them.

Card-present transactions are typically cheaper, because the risk of fraud is better mitigated. With chip cards, the fraud risk for card-present is almost entirely in physically stolen cards. (Sophisticated shops could duplicate magnetic stripes. That's a lot harder to do with chips.)

Compared to online, where all you have is the card number and the CVV, both of which are easily duplicated. And this is why tokenization is such a big deal. Tokenization means you don't store the actual card number, and the token you do store is tied directly to your merchant, so it's not widely reusable.

For low value, high volume merchants, Square's rate is extremely competitive.

I don't see how that's possible. Low value would skew it even more versus 2.9% + 30 cents, because you'll have more debit cards for low value transactions. Debit card transactions (pin or not) are cheaper post Durbin amendment.

Open to stats that say otherwise.

> I don't see how that's possible. Low value would skew it even more versus 2.9% + 30 cents

Because that’s not Square’s rate, they charge a flat 2.75%.

And a regulated debit card is around a 10 cents flat fee. No percentage.

Square's fees are pretty high unless you are processing less than say $150k a year.

From the landing page: "In-person payment processing: 2.7% + 5¢"

(may be getting A/B tested if you're seeing something different?)

Ahh, that's helpful. I went to the main pricing page. It is on the landing page, as you say...but really far down the fold.

actually, their pricing is 2.5% +10c for custom in person payments (Reader SDK) https://squareup.com/us/en/developers/reader-sdk

Square is 2.75% with no fixed fee. This is by far the best rate you can get for low amount charges and why Square enabled an entire new category of credit card merchants like fast food.

That's not that low for card present. Except for very low volume sellers. What fast food companies are using square? I've only seen square used in small shops.

For $5-10 purchases you'd cut your CC fees in half. That's a lot. Pretty much every food truck and many non-national fast food restaurants and cafes use it.

card present fees are vastly different (lower) than not present, and there is usually no fixed fee.

I work at Stripe. This isn't really competitive with Square—very different customers.

I worked in payments. Respectfully, you really shouldn't be talking on behalf of Stripe.

It's clear that a card present product, which I believe is new for Stripe, is in direct competition with Square. Even if right now it's not being marketed towards the same customer base as Square, it's clear that with a few tweaks of the app plus a change in the manufacturing process, you could easily create Square-like POS systems. Why would Stripe leave out such a huge, proven market when it's clear this product is a stepping stone towards full POS systems?

Your average pizza shop isn’t going to build a POS.

No, I meant Stripe will build it. It's not much different from what they just created.

yeah i don’t see how this isn’t obvious.

Could you clarify? I'm guessing you mean..

Square: Small/individual merchants, Etsy style physical shops, trade shows, etc.

Stripe: large retailers, grocery stores, stores with many locations etc.

What large retailers are using stripe? The fees would be too high.

I work in payments... I regret to inform you/proudly announce that you now compete with Square, just one-step removed.

I work at Square for what it’s worth..I disagree. Stripe set out to create the perfect payments service. Square set out to create the perfect physical POS. The next logical step for both companies is to spread out in the market more. We’re going to bump heads sooner or later being in the same industry.

And PayPal with their announced purchase of iZettle

I also thought it was a race to the bottom, but the rising valuations of payments companies is telling me otherwise.

All tech companies are rising. It's similar to the dotcom boom in my opinion where all tech stocks went up and everyone thought they were stock picking geniuses. Don't attribute it to anything more than this.

All it takes is for a single payment provider to start dropping rates and it will follow. Right now there hasn't been one yet, because they're know it's mutually-assured destruction. But if someone can get a handle on fraud rates, they can cut the rates down to very thin margins and work off a volume.

It is a race to the bottom.

Payment processing has become so commoditised that you need huge scale.. see the Worldpay/Vantiv merger.

What Square and Stripe are doing is adding value and simplifying the complexities around payments. They will only be able to rinse small merchants with the high fees who need the equipment also (can't afford bespoke systems).

The larger tier one retailers get payments for near zero fees as they are literally just consuming the processing and they bring their own system/hardware/management.

Wondering if this could enable use-cases for in-person subscription selling. I.e. selling SaaS/subscription content at conferences and tradeshows.

(I work at Stripe)

Yes this is one of the use cases we are excited for users to bring to market utilizing Stripe Terminal. If you also use Stripe Billing, you can help your customers initiate a subscription in-person.

Is it possible to build this use case: identifing the user using face recognition during checkout and processing payments based on his card stored in the system. Also is it possible to export the purchase data to other systems via API. So as to enable e-commerce style data analytics/customer engagement for brick and mortar retailers

Outside of being a 2FA/fraud prevention measure, what's the benefit of facial recognition here (I also imagine a lot of your customer base would be opposed to the idea of someone storing facial data for this, but I digress)? The card data should be sufficient for the purpose of linking purchases to an individual account if one is inclined to do so.

It might be slightly more convenient to be able to just walk in and out of a shop without having to insert your card anywhere, especially when you forgot to bring it with you.

On the other hand, the value of that would mostly be in the novelty, and given the current level of facial recognition technology, the hassle of disputing incorrect matches would likely make it more trouble than it's worth.

In my part of the world, most payments are through cash and very few people uses cards on day to day basis,hence face recognition.

https://www.businessinsider.my/7-eleven-facial-recognition-t...

Ah, that’s a fair point. I read it as facial recognition tying into the card stored in the system.

You mean ApplePay with FaceID?

This will be interesting as well in the consumer space.

Subscribe to lunch from your local deli for example, walk in, call your order, tap the card to verify you're the subscriber, walk out.

Or pay on demand gyms. Tap in on the card reader to either pay for a single session, or validate you're the subscriber, no extra gym card required.

"People who are really serious about software should make their own hardware." - Alan Kay

Starting a subscription in-person seems like a game-changer.

Also, attribution of in-store purchases to marketing efforts is so hard. I bet Stripe could help with this, e.g. starting with email address, or tying an online account to an in-person transaction.

I am working with a company that has small physical stores processing hundreds of payments every day... and having to rely on all the existing crappy hardware for taking payments is a pain in the a$$.. this looks like it could change the game - I'm applying for invite and see what happens next

Which is strange, because it seems to me a consumer that there's been an explosion in POS systems that offer any featureset under the sun. Are the backend systems just all wonky?

Some folks have really old hardware/software systems because the people that set them up initially (think 5-10 years ago) didnt probably feel confident with the newer vendors or maybe were incentivized to use the more older crappy systems that are more reliable but less customizable/hookable

> having to rely on all the existing crappy hardware for taking payments is a pain in the a$$

This is the same crappy hardware. Look at the page - it's Verifone terminals. It's an app based POS that communicates to standard POS card readers, so the same hardware. This is unlike Square, who built their own hardware.

Does stripe has the world best designers? Just wow.

Seems like a no-brainer. I see Square terminals everywhere, so the market seems proven. Anyone know if their pricing is comparable?

I'm not sure what Square charges, but a regular POS from your business bank would be far, far less in percentage for in-person, card-present transactions. I've seen 1.5% figures here in Europe.

So far, Stripe seems to have lowered the fixed fee from 30¢ to 5¢ and dropped 0.2% off their payment gateway fees.

I've heard Europe is more reasonable, but at least as of a few years back before Square got so popular, there was a ton of fees to having a bank-owned card reader. Typically required >$1000 deposit just to get the machine, and then a subscription in excess of $200/month, and if I remember correctly, card present transactions were around a 2.5% fee... oh, and that was typically just for Visa and Mastercard. The people I knew with these machines never sprung for the other networks because the price was so expensive. My parents run some small businesses, and credit card machine fees were one of the major line items in their monthly expenses before getting one of those cheap square phone devices back in 2012 or so. Maybe the banks have made things more competitive though now that Square has taken at least some of their business.

My previous startup has card payments from EU cards for 0.9% and 5 CZK

In EU I work for company taking 0.3% on debit cards and 0.9% for CC. Additional 0.02€ per transaction.

I use Square all day every day here at my pet store. The killer feature for Square is the consistent pricing. There are no add on fees and no clowning around. You get the same 2.75 percent on (almost) every transaction and it builds a lot of trust. There are a few cards where it goes up a little but not enough to bother me.

Stripe is well known for the same

Edit: (I work at Square)

2.5% +$.10 https://squareup.com/us/en/developers/reader-sdk

We've been participating in this trial. Happy to answer questions about it.

Can you give out any more details? Can this replace a restaurant POS system like Micros, Aloha, Digital Dining, et all? My brother is a restaurant owner and is struggling with one of those scummy legacy POS systems and I am trying to help him out of it -- even to the point of rolling an open source POS system out for him

Some ideas:

1) Flash up a QR code (and/or NFC) on the device to allow for crypto payments/buy/sell.

2) Flash up a QR code (and/or NFC) to sell custom store tokens or loyalty points.

Looks awesome, already requested an invite and can't wait to put it in practice.

The next thing that I really want from Stripe is an API for the analytics they generate. I would love to be able to display the data that they are already showing in their web application via something like their elements library. I want my customers to have a better sense of how their business is doing, without having to duplicate the work that Stripe is already doing.

I think this must exist? I know you can push the Stripe analytics data to Databox.

I'm not sure what they mean by creating a "unified experience" and I'm not sure I understand what exactly can be customized.

(I work at Stripe)

The end consumer will have a seamless experience between paying online and offline -- think buy things online, return in-store (or vice-versa). Your customers can initiate a subscription in-store that continues online with Stripe Billing. You will have all your reporting -- for both online and in-person transactions -- within the Stripe Dashboard. In terms of customization, we think our new SDKs will make it really easy for you to build the right checkout experience for your business. Also, you’ll be able to customize some aspects of the reader display (e.g., logos, color theme, splash screen, etc.).

This makes sense. Thanks for clarifying.

R.E. "unified experience", this slots so cleanly into Stripe's stable of products: all of your reporting would be available in Sigma, all of your fraud data available in Radar, and with a little magic in the SDK side, all of your customer data available (unified) in Stripe's database.

Imagine if you're running a web store, and you want to go to a conference or convention or something and start selling your stuff there. All of your tools still work, all of your reporting still works, and crucially your accounting flow remains identical.

The customisation options are probably up in the air right now - I imagine developers won't get the ability to customise the UI available on the PIN pad in the first release, but I imagine that's where they want to get to.

(I work at Stripe)

> I imagine developers won't get the ability to customise the UI available on the PIN pad in the first release

Actually, you'll be able to configure the splash screen in the first release, and will open up more hooks over time

Awesome news - can't wait for this to land in the UK

Using Verifone hardware? I'd rather just use Square. Nothing wrong with using both Square and Stripe for credit card processing.

(I work at Stripe)

We currently support two devices and will add support for more in the months to come. Please request an invite and let us know if there are particular hardware devices you are interested in us supporting.

How does the hardware work with the web SDK? I didn't think it was possible to interface with hardware from a browser.

Verifone supports HTML subset for user interface, I'd imagine that's the only thing you can customize.

Living in Russia i every time wondering why Stripe or square companies exist. We have batch of banks which doing this card acquiring, mobile phone operators who made pos terminals with internet connection and online tax services. And now i see stripe pos like back in years ago. Wtf.

Pretty interesting. I wonder if this is going to eventually turn into something comparable to Shopify's hardware POS offerings[0].

[0]: https://www.shopify.com/pos/features#hardware

I really hope they have a web based API, not requiring a native app. If they do, we would love to use them for our PWA (https://usebx.com) as a bunch of customers ask for this regularly.

It appears there's a Javascript SDK listed as available.

I think that really separates them from the rest. With everyone else, you need a native app to integrate with their terminal. Stripe offering this is really very cool!

Their js api only works for the bulkier Verizon Reader.

(I work at Stripe)

Yes -- we have both native APIs (iOS and soon Android), and a web based API (JavaScript). Our JavaScript SDK is designed for you to easily extend an existing web-based application to enable in-person payments.

Great! Any idea when you're launching in the UK?

(I work at Stripe)

We’ll be expanding support for businesses in other countries throughout 2019. If you submit an interest form for the beta on stripe.com/terminal, we’ll be sure to let you know when we’re in the UK.

Absolutely great. This industry(card processing - esp. in person) is mired by a ton of ambiguity and it would take a company like Stripe to open it up to innovation. Cannot wait to see what comes out of this from the ecosystem.

I couldn't agree more.

Up to now it has been near impossible for third party solutions to end up on terminals. I've seen these new Verifone terminals and they are awesome. I can't wait to see what comes of these.

We have developed a POS for salons and spa and have been working with various in-store payment solutions - including Square. None of the existing solutions meet our needs.

I have applied for an invite - can't wait to get access.

(I work at Stripe)

Excited to review your application! We have a lot of users building point of sale applications for booking and scheduling -- including AtVenu, Mindbody, Universe, Zenoti.

How likely are startups who are still building out a platform get approved for beta access? And should startups add that info in the beta request form?

(I work at Stripe)

Yep, please tell us as much possible in the form. The Beta is open to all business in the US.

Can someone clarify what the use-case is for this new functionality? I've regularly used POS stripe devices based on iPads, is the idea this allows Stripe to be used on proprietary POS devices?

(I work at Stripe)

Stripe Terminal gives developers the flexibility to build custom point of sale applications that use the Stripe payments platform, with EMV-ready hardware. The product will be work across the entire suite of Stripe products (Billing, Radar, Sigma etc.), and we’ll provide native SDKs across several languages.

Side note on the design - Absolutely love how well they do with keeping depth on flat design elements. Stripe design team always does such good work.

So, no need to worry about PCI compliance, right?

Well you're not holding credit card details, so no there should be no need to worry about PCI.

I was hoping to hear that from Stripe. The word PCI is nowhere on the headline page.

(I work at Stripe)

Apologies for the delay! Stripe Terminal works with our pre-certified card readers. This ensures that your transactions are secured by our end-to-end encryption and that your readers are always up-to-date via our remote management tools. We’ll share even more information about how we help reduce your PCI scope with Stripe Terminal when we launch full docs later this week!

I am naive in this space, could some explain what is keeping the duopoly of Visa/MasterCard going?

While I haven't done a lot of research in this area, I would suspect launching a new credit card company is just a massive amount of work, very different from launching a new web app. You're obviously going to be subject to a lot of banking and capital regulations; you're either going to have to be starting an actual bank or you're going to have to partner with (at least) one, for instance. Then you have to figure out how to get businesses to start accepting you. "This card works everywhere that happens to have a Square-based point of sale system" is not going to get you to be the next Visa. It's not even going to get you to be the next Discover.

While Simple was trying to launch a new bank, not a CC processor, their experience is probably informative here -- they're successful by some measures, but they failed at disrupting the industry the way they wanted to. (I'm a Simple customer and have been on the edge of switching back to a conventional bank for over a year; it's mostly just inertia that keeps me there.)

> While I haven't done a lot of research in this area, I would suspect launching a new credit card company is just a massive amount of work

there's a Futurama episode that has a snarky commentary on this - s01e06:

fry: "Do you take Visa?" clerk: "Visa hasn't existed for five hundred years." fry: "American Express?" clerk: "Six hundred years." fry: "Discover Card?" clerk: "Hmm...sorry, we don't take Discover."

I wonder if Stripe or Venmo will be the first to get QR payments going in the US.

CurrentC was created a merchant consortium to push barcode payments (as seen at CVS, I think?) and didn't get anywhere. It's not clear to me why you'd want to use a barcode rather than NFC a la Apple Pay.

https://en.wikipedia.org/wiki/Merchant_Customer_Exchange

The likelihood of QR payments taking off in North America is slim. People love credit cards. They offer way too much.

This model worked in China because people went from using cash to mobile.

We've been using credit cards forever and there's literally no incentive for me to use an app like Venmo/Square Cash over my Chase credit card which gives me points/miles

It is simple as this and I have seen the research.

The west loves tapping and the east loves scanning (AliPay / WePay)

anyone know if it will support pre-auth .. I own a nightclub and this would be clutch. I saw they have a preauth endpoint. But, can I preauth for $1 and capture $100

Of course Canada left out again. lol

Hiya, I work at Stripe. We want to make sure that Stripe Terminal supports the major in-person payment methods in market. As you know, Interac is really important in Canada, and we're working on adding support for that. Stay tuned!

Competition is good I guess...

Can it be also hacked?

Reply on news.ycombinator.com