(I work for Stripe, and since many of my colleagues are currently sleeping, taking the liberty of saying this on their behalf.)
We are looking into things.
As someone who ran businesses for a long time myself, it would be very alarming to me if customers felt they were being ignored while repeatedly talking to people who work for us. If that ever happens, please bring in literally anyone at Stripe, inclusive of other Patrick. Our email addresses are often available through a quick search of HN; mine is (predictably) patio11@stripe.com if you need it.
We don't comment publicly on individual customers. This is an important part of customer privacy and we're serious about it.
Sometimes situations which result in external complaints are the result of a process failure, and in those cases we try to fix the instant case and improve processes in the future.
Sometimes they are the result of a process operating correctly but coming to a result which someone did not enjoy. The record available to the process is often more detailed than the record available to the public Internet, and may include extremely relevant context.
Corporate process improvement is not an excuse for treating people badly. An individual human who treats people badly is not allowed to say "well I deal with a lot of people and I need algorithms and sometimes my algorithms don't work so I'm doing my best to improve them". Corporations should not get a pass because they need cheap processes that "scale" to maintain their profit margins.
This is the perpetual moral hazard that we constantly see at all big tech firms. These firms promised fantastical profit margins, but only on the condition of scaling at very low cost. What doesn't scale at very low cost? Customer service. That's why you're all so awful at it.
And don't come back with your satisfaction metrics from customers you haven't screwed. Your job is to do right by the customers who aren't convenient for you. Just like it's an insurance company's job to pay out to the tiny minority of claimants. It doesn't matter one bit if the company has a great web site or API that everyone loves. The true moral nature of your company is revealed by how you treat the customers who need you to do right by them.
Perfect
Disagree but a morally commendable position.
Also: Stripe publicly emphasized the possibility that their customer might be a dirty rotten liar.
I feel it is only fair to point out that Stripe might be a dirty rotten liar, too.
It turns out the poster did, in fact, omit extremely material info from their original post. They sold a used company van via credit card via their stripe account: https://news.ycombinator.com/item?id=32264886
This is not something the business sells, and a huge chargeback risk
Isn't the obviously correct behavior by Stripe to suspend that transaction, without interrupting others? Clearly they are able to identify transactions on a case-by-case basis.
I read their account. They only complain of Stripe holding "the" money, from the car. Did they say Stripe locked their account or did any other restrictions?
https://www.reddit.com/r/smallbusiness/comments/wa1zob/dont_...
I read a few of their comments their complain seems to be about the $3000 and needing to wait 120 days to get it.
They also say you can't call, but at least for me I'm able to call Stripe here: https://support.stripe.com/?contact=true
I don't see any phone number there.
If you have an account, you click "call me" and they call you
OP said the options to "call me" is greyed out: https://old.reddit.com/r/tifu/comments/wa230m/tifu_by_using_...
First, that seems to be what happened to OP. They suspended the transaction and are holding funds associated with that transaction.
Second, I posit that having even one (confirmed) suspicious or fraudulent transaction increases the likelihood of having others.
From elsewhere in this page:
> I used Stripe for about a year to run a small cell phone store in Denver, CO area. [1]
> We sold a cheap van in the company name. [2]
Yeah, that explains how they arrived in Stripe jail.
[1] https://news.ycombinator.com/item?id=32261868#32264335
[2] https://www.reddit.com/r/smallbusiness/comments/wa1zob/comme...
Should’ve used Zelle or cash. At least with your bank, you can apply regulatory pressures. With Stripe, you have little recourse.
I think this negativity is uncalled for. They said that they do their best, sometimes there are mistakes on their end that they try to fix, and sometimes the person complaining is lying or leaving out details that would explain the situation. If you've ever dealt with people, you would know that they often "forget" things or downplay things that would make them look bad.
I think this criticism is called for. The customer tried asking them nicely and it sounds like they did not try very hard to fix it -- which is an experience I can empathize with 100 times over. If you've ever dealt with companies, you know that they often have policies and procedures that intentionally screw customers and they only back off if it blows up publicly. Because they can. Our accountability mechanisms suck.
I don't think you - or many of the people commenting here - read through the thread. The customer personally sold a "cheap" van and accepted a credit card as a payment, which is super, super sketchy. Of course Stripe flagged it. They were smart to do so.
You're bringing negativity as well. Just applying it to the smaller, weaker party.
When so much money is at stake, the negativity is called for. Stripe is not neutral here. They have leverage and power over their merchants in these situations.
I read OP’s reply as a generic criticism about companies that scale beyond their ability to provide sufficient customer support, not a specific criticism of patio11. Silicon Valley is full of companies that scale beyond their competence and ability to operate properly: it’s practically a requirement of the “growth at all costs” mentality. I would argue that if you cannot handle dealing with customers without AI and automation with many false positives, then you have no business scaling further until you can. I don’t know if this is the particular case for Stripe since I’m not a customer, but we can all see this behavior across the industry.
That's an incredibly charitable interpretation in favour of Stripe. The aggrieved party has been emailing their support repeatedly only when this gained traction on both Reddit and Hackernews are they actually replying. How is this acceptable?
I don't see how you can assume bad faith on the user. Have you ever dealt with corporations? If so, you would know that they often lie and downplay things that make them look bad.
> Corporate process improvement is not an excuse for treating people badly. An individual human who treats people badly is not allowed to say "well I deal with a lot of people and I need algorithms and sometimes my algorithms don't work so I'm doing my best to improve them".
I don't think you're disagreeing with patio11 here. To wit:
> Sometimes situations which result in external complaints are the result of a process failure, and *in those cases we try to fix the instant case and improve processes in the future*.
I read that as: they prioritize satisfying the immediate, dissatisfied customer first and address process improvement second, which is exactly the priority that you're suggesting. Especially combined with patio11's other suggestion of personally contacting the upper management of the company.
> Our email addresses are often available through a quick search of HN
Your customer service strategy is really "go to a small internet forum that only some startup nerds know about and find my email address"?
Makes you wonder who the audience for the original comment really is.
Yes, in Stripe’s case. That’s the only way I’ve ever had a big issue successfully resolved. They empower their customer service team with over a dozen ways to say “sorry the system won’t let me solve this” or “Your issue may be unique to you, but since my solution for somebody else’s different issue wouldn’t work for you, I have no power to do anything”.
I doubt it's the general strategy. Patrick might be assuming that frequenters of HN are less likely to be bad actors and people that legitimately are slipping through the cracks.
Customers shouldn't have to bother random employees to get support on critical issues. In general I think Stripe does a pretty good job with support compared to other big tech firms - the inability to reach humans is problem endemic throughout the industry.
It's a shame because a good customer support interaction can create a serious evangelist (hard to quantify but extremely valuable). Maybe Stripe knows this because they put a lot of effort into appeasing the HN community (many developers/entrepreneurs who make processor purchasing decisions). I wish they'd apply this attitude to all their customers.
Maybe if the company had some way, any way at all, to reach an actual human being, this wouldn't be a constantly repeated problem on HN and elsewhere.
As I have observed in other contexts ( https://twitter.com/patio11/status/1463749947858165761 ), many people who say "Ugh, I cannot reach a real human" have in fact spoken to several humans, repeatedly and at length, and are imprecisely stating their desire to speak to a person who would agree with their view of the facts and swiftly enact the resolution that they prefer.
We offer (human-powered, somewhat obviously?) support via a variety of methods 24/7; time for us to reply to email is generally a few minutes. Could we do this better? Yes, and I hope we continue doing it better; if anyone ever has an interaction where we don't meet the bar please flag to us.
In this specific case, the OP claims that the call and chat options are greyed out, leaving only e-mail and social media. E-mail, they claim, is bot responses only.
While I understand you can't comment on the specifics of this case, can you speak generally about whether or not disabling the phone/chat support buttons is a part of some standard process?
Speaking generally, chat and phone support are designed to provide immediate round-the-clock service to a wide range of issues but not literally every type of issue that a financial institution has to be able to deal with.
Sometimes, a financial institution might only be able, due to staffing or other considerations, to deal with you asynchronously in writing, perhaps from e.g. a group of professionals who are not as numerous as front-line user operations.
This is a roundabout non-answer that doesn't explain why chat and phone support were/are greyed out.
Are you just trying to say "we're so short staffed, there's no one to answer your messages, let alone your calls?"
What he is saying is that in the instance of a fraud concern, they don’t have people who can talk about it available 24/7.
They don't have it 24/7. That's part of the answer.
But that kind of phrasing leaves a big gap.
Do they offer it during some hours? Which hours? Or do they offer it never?
There's a big difference between "call and chat are only extended beyond business hours for certain issues" and "call and chat are only ever for certain issues".
The response here could mean either.
On this page where it pops up a support window, https://support.stripe.com/?contact=true (you might need to be logged in to see it) there's a dropdown with various choices, and the Chat and Call buttons get enabled or disabled depending on what you choose from the dropdown. In the default state with no option chosen, all 3 buttons are grayed out, but a lot of the options seem to enable all 3 buttons.
> imprecisely stating their desire to speak to a person who would agree with their view of the facts and swiftly enact the resolution that they prefer.
Wow, one reply in and we're already seeing open contempt for this type of troublesome customer. What nerve, expecting a payment processor to promptly process payments and make funds accessible to the account holder.
This reply is going to look monumentally bad if the customer ends up being right about how they were treated.
The Reddit thread shows that OP has interacted with Stripe through Twitter DMs at a minimum.
Yes, this happened in the thread, after other people suggested it when he asked for help connecting with a human. So if the purpose of your comment is to support the idea that, many people who say "Ugh, I cannot reach a real human" have in fact spoken to several humans, repeatedly and at length, then you do not have enough information to support that here.
The statement that he never talked to a human could be true at the point in time when he first made the claim in his reddit post.
> This reply is going to look monumentally bad if the customer ends up being right about how they were treated.
And will you say the opposite when the opposite is true? Other comments mention facts about how the OP tried to do something outside of normal and the response is expected for "abnormal" behavior on an account.
Anger against companies is one thing... but will you defend them when it turns out the company is in the right or is this only a one sided comment?
This comment should make every Stripe customer run.
You've build up an incredible amount of kudos with your comments and blog posts over the last decade+ here and are willing to burn it for a corporation in a moment? Earning trust takes a long time, losing trust only takes the blink of an eye.
The guy tried to use a cellphone shop account to sell a used van via credit card. If I was Stripe I'd be holding the money until the chargeback period was over too.
They should not accept the payment and then keep the money. Why not reject it in the first place (too large of an amount)? $3000 reasons.
They aren't "keeping the money," it is in escrow until the dispute period ends.
And who owns this "escrow"? I bet it's Stripe. That money is sitting in their accounts collecting interest that they definitely want (spoiler, they make money off keeping your money in your accounts not just off processing fees).
Is selling a van universally against the terms of service? Especially a cellphone shop van?
If it was just bad communication about what this specific business uses the card for, it seems like that should be resolvable without waiting out the chargeback period.
When you get a merchant account, part of the application process is describing what your business is, what it sells, and what your expected sales volume is. These metrics are used for fraud detection and risk management.
Using a merchant account for a purpose which you haven't disclosed to the processor -- like a cell phone store selling a truck -- is a violation of the merchant agreement, and is often grounds for account closure.
I know that. You avoided addressing either of my points though.
1. Is selling trucks against the terms in general?
2. This is a cell phone store selling a cell phone store truck. It's part of the business, not something unrelated.
If they had thought of it, they probably would have made trucks or no trucks a part of the processing agreement up front. They thought it was a reasonable use, and it might have even been a reasonable use! So why can't this be resolved any sooner or with better communication?
> 1. Is selling trucks against the terms in general?
Probably not, but I can't rule it out. It's certainly an unusual thing to accept a credit card payment for, and the fact that vehicles are titled property may complicate matters.
> 2. This is a cell phone store selling a cell phone store truck. It's part of the business, not something unrelated.
No, it isn't. The business of a cell phone store is, generally speaking, going to be selling cell phones and cell phone accessories, and its merchant account will have been obtained with the understanding that it would only be used for charges related to those products. It isn't in the business of selling cash registers or furniture or trucks, even if it may happen to possess those things as a business; if it wants to sell those things, it needs to accept payment through some other means.
> generally speaking
> its merchant account will have been obtained with the understanding
> if it wants to sell those things, it needs to accept payment through some other means
So it's not part of the general business, but it is the business doing the sale, and such a thing was not part of the merchant account negotiation.
...I don't see a single thing you actually disagree with me over?
I can attest to this.
The small company I was working for got card tested (scammers use hundreds of stolen cards to see if they work). I called customer support and talked with a human on a Friday night at 8pm for 15 minutes about our options. 2-3 days later, I chatted with someone from the fraud team.
I was surprised by the high touch service even though didn't have a lot of volume (new account w/ less than $10k of transactions).
Agree. I've been a Stripe customer pretty much since they launched, I'm small potatoes but they have always been very responsive and with smart support staff.
"Am I so out of touch? No, it's the customers who are wrong."
I’m not a Stripe customer, but I’ve often ended calls with other companies’ human tech support thinking “wow, I wish I could talk to a human.” Not because they didn’t take my side, but because I might as well have been talking to a chatbot. Many companies’ tier 1 support can’t go off script, can’t offer creative solutions that are even slightly off the rails, can’t make exceptions, can’t apply human reason or good judgment, can’t update fields in their screens that are read-only… they are basically IVR systems that eat lunch. So I can emphasize with the feeling that customers are always interacting with an API.
Yeah, I'm also not a stripe customer, but it's not unusual for me to email a company, and get back a canned response to a tangentially related question I didn't ask. If I then reply and try to clarify what I meant, I get back exactly the same canned response.
I don't know whether it's a chatbot spitting out replies based on keyword matching, or whether it's a human with a "responses per hour" performance target which doesn't give them time to read the emails they're replying to, but either way it's very different from the "talking to a human" experience I'm used to from daily life.
(And it's not a case of a correct process coming to a result I don't enjoy, except in a very narrow sense. "I understand your request and the answer is no" would be an improvement over what I often get)
This happened to me with Tinder, when I begged them for 2 weeks before expiration to let me renew my grandfathered Platinum membership, since my card wasn't showing up. Their canned response kept saying the account will auto-renew, no matter how I worded "no it won't, because you don't seem to have my card on file, even though my card hasn't changed".
Then, when the account didn't auto-renew because they didn't have any card on file, their canned response switched to "create a new Platinum account" (with much worse benefits).
I requested that this be the same grandfathered type of Platinum account that I lost.
Their canned response was that I should create a new Platinum account.
I said okay, but I want the same benefits I lost.
Their canned response was that I should create a new Platinum account.
I created a new Platinum account. It had much worse benefits.
Approximately one week later, I was permanently banned from Tinder for violating community standards without explanation (after having been a nice member of the community for the past year on Platinum and continuing to be one).
So in your first post you state:
>We don't comment publicly on individual customers. This is an important part of customer privacy and we're serious about it.
But then you heavily imply that the user ("Many people who say...") is a liar? Bad look.
What does "human-powered support" even mean? Is this weasel speak for a chatbot or do you mean actual people replying to support claims?
> We offer (human-powered, somewhat obviously?) support via a variety of methods 24/7
> time for us to reply to email is generally a few minutes
These two assertions are conflicting. A few minutes is not nearly enough time for a human to understand and address a support request. So like most companies, it sounds like Stripe is actually running automated support behind a fleshy humanesque interface. Don't delude yourselves into thinking that this counts as "human-powered" support.
Have you ever worked in a support role? A few minutes is plenty for >95% of support requests. Those last few percent can certainly be a doozy though.
Tech support, a long time ago, in which the quick calls were extremely rare. Diagnosing a problem on the customers end or on our end, each took a lot of time.
I've also wasted hours of my life on hold with customer service "people" who, in the best case, take 5 minutes to hear my situation, 15 minutes to check into it, 5 more minutes to apparently not do the thing they said they were going to do, and 0 seconds to forget about all these details after hanging up, meaning I had to go through the same pattern all again the next call. Even in the cases where I've directed reps to look at previous tickets for history, half the time they still can't find it.
I'm sure that there are many people that use support instead of clicking around to the right option on the website, even though that seems abhorrent waste of time to anyone with half a clue. But companies optimizing for this are doing a disservice to everyone that is only using higher touch methods because they're trying to solve uncommon situations.
I can believe that "modern" support only takes a few minutes per the request, because each individual request is itself just a cog in the larger machine. When I spend an hour on the phone, bounced back and forth between 3-4 departments, how long was each of those conversations? Like so many things these days, the problem is with the overall constructive behavior rather than the myopic metrics that have been over optimized.
And don't get me started on those useless freeform voice prompts. Don't make the customer distill their problem down into a curt description, because once again, they're likely calling with a weird problem. Just give me the numeric options that mirror your business structure and I'll pick the most appropriate one, thank you.
Honestly it sounds like you use call or email support as a last resort, which is how I use it too. I poke around on my own, look through the documentation, check out the website, etc.
Many, many, many people do not do that. Ask any retail business what the #1 and #2 calls are and the answer is always #1: "Are you open right now?" and #2: "What are your hours?". I suspect a big chunk of Stripe's email requests are similarly trivial - "where can I see my $whatever report", "can I take amex?", "how much will I pay for a $.95 transaction", etc. that a competent support person can answer in their sleep. Now if we're getting into complicated technical issues, I'm sure those should be taking more time.
I share your disdain for those "just say what your issue is", although the trick to that is usually to say "cancel" and get immediately transferred to someone who will be expected to keep your business.
A business is a structure of people, and my intuition is to achieve healthy dynamics, people should treat it like this kind of parasocial relationship. If you fuck up, I should forgive you as long as you're not an asshole about it. However, it seems like a recurring theme, that modern tech companies don't actually provide support. If that's not in line with the economics of the situation, maybe I'm wrong to hold it against the company itself.
But maybe the cost of this model is hidden from the consumer, and I'd rather pay a bit to have reliability and fairness. I don't know how much I'd have to pay, but I'm willing to guess it's palatable, or at least in the neighbourhood. If we are being tricked by these hidden costs, it's better (if you can) to move on, there are people trying to compete even if it's a difficult industry to break into. Try not to get locked in. If I were running a small business, I'd definitely be discounting crypto payments and putting the extra work into handling whatever regulatory complications arise, vs yelling at the walls that are the infinitely scaling tech companies
I only have one recurring bug bear that seems trivial for Stripe to solve:
Is there a reason you can't get ARN/STAN refund codes via the Stripe API?
I have never been able to get an answer to that.
Don't worry. I made your correspondence public concerning my account from last year so people can see how you handle complaints.
> I work for Stripe
Bit of an understatement, right?
He is not Patrick Collison.
https://www.kalzumeus.com/
Wrong Patrick: kalzumeus is Patrick McKenzie. Patrick Collison's site is patrickcollison.com, and he's 'pc' on Hacker News.
thats why he shared mckenzie's website; he's the commenter
Customers are publicly asking for your answers for your what in their opinion is an ill conceived and/or villainous process. Not answering claiming that it violates your privacy process from which they specifically exempted themselves could also be considered even more ill conceived/villainous.
They are waiting for your answers, it is a public matter as per your clients repeated requests, it affects more people, money is involved, we are all waiting for your resolution and future mitigation actions. Or rebuttal ofcourse.
If I was a customer and Stripe (or any other company) started releasing private correspondence and financial details publicly because someone on the internet claiming to be me complained about the company, and the company took that as implicitly waiving any privacy obligations, I'd be pretty upset. If they said that the prurient interest of their other customers in the details was part of their decision I'd be even more upset!
> Sometimes they are the result of a process operating correctly but coming to a result which someone did not enjoy
So much to unpack here but this is extremely patronizing ("we're not wrong, you just didn't like what we did). "process operating correctly" refers to holding a merchant's money with no reason communicated, and no way for them to get in touch with the company?