Service Providers, Business Sustainability, And Society

I’m not sure how many people think consciously about the business plans of the service providers that they start using, or at least relying upon. I don’t do it terribly often, but I do sometimes, and I thought I would share some words on the topic, because I do think that the world would be a better place if we did think about the effect of our choices to the wide world.

I have for instance wondered aloud, over the years, on Twitter and even on this blog, about the fintech company Curve. Their services are actually interesting and valid… but they do not offer any interesting value at a premium — I would never pay for their “Curve Metal” tiers, because it just doesn’t add up. I couldn’t just figure out how they were expecting to keep running, given I expect the majority of their “customers” are consumer end-users that would follow the same procedure I followed: sign-up for the service, using the £5 offer, use it for the 90 days of free cashback, possibly use it a time or two while traveling, and otherwise just keeping it as a backup card. Before the pandemic I also wrote how they were giving out more free money, but more recently they decided to start crowdfunding. And they became very loud when they did. Which is what reminded me I had a half-drafted post on the topic (this post), which I should probably resurrect (which I did, since you’re reading this now).

Before going back to Curve and their crowdfunding, I want to point at two sayings that you’ll keep finding around you, when you discuss businesses, products, Internet, and privacy: «If it sounds too good to be true, it probably is» and «If you are not paying for it, you’re not the customer, you’re the product». These are good places to start a discussion, although I think there is a lot of nuance that is lost when trying to (over) simplify with these.

Full disclosure: I work for a big company that, primarily, offers services to end users free with ads, though the product I personally work on not only does not relate to ads, but does not even have ads in it. And while my previous employer was another big company that is part of the “AdTech” business (and in there I did work on ads systems for a long while), I have discussed ads in the past, and I even ran ads on this blog (back in the days before stable employment), so you can imagine that what I’m going to be writing about is my personal opinion and does not represent that of my current, past, or future employers.

So why do I think it’s important to figure out the sustainability of service providers? Well, because it becomes a problem for the whole of society when a fraudulent or scammy provider gets a certain about of market share, even if sometimes not evenly and not in a way that most people would be able to connect together. For instance you can take the examples of Enron, Bernie Madoff, Theranos, or Wirecard — organizations that promised too-good-to-be-true services and profits, and ended up bust, with different blast radiuses. For the last one we don’t even quite know yet what the blast radius will be once things settle down: the German financial services environment is likely going to be reshaped by last year’s scandal, and so it appears will be EY.

Though this is clearly not limited to financial services (VPN providers seem to be pretty much in the same position), it does look like the likes of Curve and Revolut are easily the most visible cases where a company apparently lacked a sustainable business plan, and decided to turn to a crowdfunding campaign — Curve just had one this past May, and it was so noisy that I know a couple of people who went on to find a way to delete their account (and the app) simply because they got tired of their pushy notifications (not a typo).

Now, that might sound not too bad — after all, crowdfunding for the most part just means someone is willingly going to pay to subsidise other people’s “free money”. But the next thing that Curve did after that was to increase referral bonuses for new users to £20, from the previous £5, and that smells to me even more fishy — because that sounds like trying to bring in a mass of users hoping that enough of them can’t figure out that the premium options of Curve are not worth the money.

On the other side of the tracks, Revolut has been pushing more and more for cryptocurrencies, which I’m not going to even pretend is a neutral thing. I care enough about the environment that their consumption alone makes me angry, but even more so, I find that the amount of scams related to cryptocurrencies at this point are wide enough to show that the whole concept is hostile to society. I do not support nor recommend Revolut to new users unless they live in countries like Ireland where there is no other option – in London, using Revolut feels like subsidizing scammers by lending respectability to cryptocurrencies.

But at this point, neither of those appear to have reached the full consumer scams, so it should be fine, right?

Well, let’s take a different example with the VPN market. I have complained over on Twitter a few times how I blame us geeks for the amount of VPN scams that are out there. Privacy maximalists tend to scare people with the idea that your ISP, the Starbucks, or the airport lounge you’re using can see everything you do — and while it is definitely the case that there’s a lot more data going around than you may think, ads such as those ExpressVPN pays the otherwise excellent No Such Thing As A Fish podcast to air, that suggests that your ISP would be able to tell what you’re Googling are not just falsehoods, but proper FUD.

But even accepting that ExpressVPN has no ulterior motive and are totally legit – I have no idea about that, I only used them before while in China – and leaving aside the fact that VPNs have huge targets painted on their backs, there’s still the matter that you need to trust your VPN provider. Which may or may not be more trustworthy than your home ISP — the two of of them having pretty much the same power. Most of the review websites seem to be talking more about commissions than trust, because the worst part is that there is nothing that allows you to verify their statement that they are not logging your traffic in the same way they keep insisting your ISP is doing.

So how do you trust a VPN provider? Well, for a start you may want to start considering who their founders are and how they get their money. And you’d be surprised how many dots you can connect this way. For instance, last year CNET wrote about Kape Technologies, a company that bought a Romanian VPN service called CyberGhots. In that article, they also noted that in addition to CyberGhost, the same parent company bought two more VPNs:

After buying CyberGhost, Kape then bought VPN ZenMate in 2018 and more recently Private Internet Access, a US-based VPN, in a move which Erlichman said in a press release would allow Kape to “aggressively expand our footprint in North America.”

Now, the problem is less about a single parent company owning multiple VPN services as they were different brands — this happen all the time in many other fields. Just look at the relationship between Tesco Mobile and O2, or banks such as Halifax and Lloyds. But the rest of the article does make for a good build up for why the whole situation is a bit suspect.

But more importantly, you may have heard of Private Internet Access before — they are the company that started heavily sponsoring Freenode a few years ago. And if you have been paying attention to Free Software projects’ communication in the past few months, you probably know by now that Freenode is a trash fire now. So given those connections, would you trust anything that has connections to these organizations and people? I clearly wouldn’t.

This same problem with trust and business sense applies to other businesses. With the exception of B Corporations, most companies out there are intended to make money. If nothing else, they need to make money so that they can pay the wages of the people working there. So I don’t generally trust companies that appear to be giving everything away — and rather prefer those that, if they are making money with ads, say so out right.

In the case of Fintech services — Wise (formerly known as TransferWise) is my example of choice for a company that is transparent when it comes to the cost associated with their services, and makes a good case for why they charge you, and how much so. I really wish more of them did the same because it would make it easier for people to choose how much trust to put in a company. Unfortunately it appears that the current trend in the market is to push as much grown as possible for companies to grab a captive audience before turning on the monetization screw.

Important note: this blog post was written before Wise announced they intend to go public (it was previously rumored, but I didn’t spot that). I guess I should now disclose that I will most likely consider buying some stock of the company, though probably not on the IPO day. We shall see. As I said, I do like their business sense.

Going back to a moment to that «if you’re not paying for it, you’re the product» as well — well, I don’t agree in full, but this is something that people do need to be keyed in to look out for. In particular, I don’t think that ad-supported businesses should disappear, and that everything should be hidden behind a paywall, because I do think that having wider access to information without making it costly is a good thing. But also I do think that there are services that are often crossing the line into being creepily interested in your data rather than “trade it” for useful information.

But I also think the scrutiny is often placed more on the big, established companies rather than the “scrappy” startups, or the more consulting-like companies. Heck, a few of you reading this are probably already ready to complain that both my current and past employers are seen as data hungry — but I can tell you that both companies, at least during my tenure, would never have someone state on a stage that collecting data from IoT sensors and just throwing it to a ML pipeline to gather unexpected insights, as it would go against every one of the privacy and data handling trainings and commitments…

And yet John Roese from Dell EMC stated that in his opening thoughts for LISA 16 (go to minute 44 in the open access video) in what sounds terribly like an advice to startups. To be honest, that’s not the only cringey thing in that opening talk — from a technical point of view, his insisting that persistent memory means you can’t just reboot a computer to reset the state of memory (as if re-loading the data in memory from scratch wouldn’t happen on request, whether this is persistent or not) is probably a worst phrase.

What I’m trying to say is that you need to be sure who your friends are, and it’s not as easy as to expect that all small players are ethical and all big ones are not. And asking yourself “how are they making money, if at all?” is not just allowed — it should sometimes be considered a necessity.

Curve is giving “free” money away, again.

About a year and a half ago, I reviewed the Curve debit card, and I went back talking about it when talking about foreign transaction fees. If you don’t want to go back and read the whole set of text, I’ll give a very brief description: Curve is a proxy-card, that allows you to connect a bunch of other debit and credit cards, and to decide when you pay (or, critically, shortly after) which card you want to charge your expense to. It includes a few features such as some amount of free (or cheaper) transaction fee spend, but the “proxy” nature of the card is the selling (or not-really-selling as I’ll explain later) point.

In my previous posts, I have made two main points about Curve: the first is that they give away for free most of the useful features of the service, and the other that they make no real sense in the UK, as one of the possibly biggest selling points (the foreign transaction free nullification) is vastly irrelevant: most high street banks have some offering with no foreign transaction fee, and I still venture that Santander is the best UK offering for globetrotters who need to use their card in many different currencies. Despite my fairy pessimistic view of Curve’s business plans, it seems like the management is taking a different view — even FT Alphaville wrote about their marketing campaign.

Speaking of pessimistic view — I am a bit skeptical about their marketing of “100 Cards in One”. While it would be a great feature a few years ago, in 2019 most of my spending goes through my phone, with Google Pay. While I have half a dozen separate cards, most of them are compatible with Google Pay, so I don’t carry them with me. The ironic exception being my company card. As it turns out, this is something that Curve can help with: it now supports Google Pay, and on a business trip I can proxy my expenses to the company card.

Now, one claim on their website that appears to mostly hold true is «Your gateway to money for nothing.» Because they do appear to run lots of promotions that give you free money. Indeed, in the past week I received two email from Curve: one to announce that they would give me £5 if I just used my card at all (which I did, just to see if they pay up), and another to tell me that they are giving a “Christmas gift” for all their users to select three new retailers to get their 1% cashback from. All of this for a “Curve Blue”, which is their totally free tier.

Speaking of the 1% cashback, when I signed up last year, Amazon was not one of the options, or I would have taken it. It was this time, so I did that, under the impression that one way or another I do end up buying stuff off them often enough, and in the next three months I may get some value out of it. Despite this, their paid offer is still pointless: they charge you £9.99 a month, and to cover that on cashback alone you would have to spend £1000 a month from those three retailers. And no, I don’t think the Travel/Gadget insurances that they peddle with the offer mean anything else — there’s a doubling of how much free cash you can get from ATMs, but they appear to have closed the loophole that allowed you to withdraw cash and get loyalty points, or cashback, from a credit card, without incurring in cash handling fees.

So yeah, it looks like they do give money for nothing. Well, for some profiling data I guess. The obvious question is where that money comes from, given that the free offering is just compelling enough, and their paid offering are… in one word, overpriced. As I said in the previous post on foreign transaction fees, Santander offers their All in One Credit Card for £3 a month, and is also 0% foreign transaction fees and comes with a 0.5% cashback on all purchases; recovering that monthly fee requires “only” £600/month spend across any vendor (and not just three), and if you spend more you can probably pay for the travel/gadget insurance separately. And since it’s issued as a World MasterCard (rather than a Debit MasterCard), it also allows you to use some of the available perks worldwide, including some airports’ priority lane at security (as it turns out, that includes Venice Airport, which is very handy since that’s where we fly in and out for to see my family.)

And if you want to compare with the Curve Metal offering at £14.99 a month, well, Santander offers a World Elite Mastercard at the same price point, which comes with the same 0.5% cashback (although capped to nullify the monthly fee.) Despite not coming with the insurances (which again I don’t find particularly compelling, it does have a discount for Santander’s own offering. And it provides LoungeKey access just as well, except that you don’t have to pay the £20 per person entry fee. Being a World Elite card, it also comes with a bunch of other perks, including a Boingo subscription (not particularly compelling to me either, but worth noting.)

Anyway, if you are yet to make your Christmas purchases, and are interested in getting some more extra cashback with Curve, you can download the app from their website and if you want you can sign up with the code BG2G3 to get another £5 out of the magic free money card (and give me the same.)

Revolut, as of October 2019

A few months ago I wrote a not-so-short comparison of a few FinTech services with offerings from high street banks in the UK — and I would note again, that the comparison does not hold up in Ireland, so it’s definitely biased, but I would uphold it for good reason. I think it might be time to do a bit more dusting over it.

The first service I should get back to talk about is Revolut, which I first praised and more recently complained about. As I said in a number of previous posts, my reasons to keep using Revolut for day-to-day transactions have pretty much disappeared: my Santander credit card gives me 0.5% cashback on all transactions, and no foreign transaction fee, why would I use Revolut? Virtual cards, and rotating-number cards are interesting and have their use, but honestly, I can’t be bothered unless it’s for very shady operations where I don’t trust giving my credit card, but those are pretty much corner cases.

Revolut has been running multiple advertising campaign throughout the London Tube, the most recent one promising three Tube trips free if you pay with Revolut. I could probably do that, next week, maybe, if I paid enough attention — I don’t use monthly tickets, so I can change card any Monday as long as I use it until the same until Sunday to cover the 7-day cap. But I had bad history with using Revolut on the TfL network before, although admittedly that was when I was landing from Dublin, and the location-based security tripped.

Update 2019-10-07: turns out I cannot actually use their TfL offer because it relies on Google Pay (which with Revolut I found already too unreliable to use for commuting) and only works if you have a Visa-issued card. My card is MasterCard-issued still.

If you check the news, the FT reported just this week how Revolut expects to reach “viability” despite continuing to lose money. This is likely because, as I pointed in my complain-post, Revolut makes perfect sense as long as you’re not paying anything for it. The only reason to sign up for any Premium or Metal tier in London (where most of their advertising budget appears to be spent, from what I read from news) is if you don’t understand the services available from the high street, or if you want to subsidize the free tier for everyone else. Funnily enough, FT Alphaville reported on the same day of the staff cashing it in.

I had to use Revolut only once in the past few months, and that was a couple of days ago. My sister asked me if I could send her some money for her to use the card, as her debit cards expired and she was trying to buy something — remember Italy does not have “faster payments” so inter-bank transfers are not instantaneous. It should be a simple operation: top-up £50, send £50 to my sister, she can convert to € and spend it.

Topping up worked like a charm. But sending the money didn’t: in addition to confirming my fingerprint, the app said it would send me an email, and to check the email from the same device to confirm the operation. The email can be re-sent only after one minute, but (as often) it recommends you to check your Junk or Spam folder too. The email never arrived. I don’t mean within a minute. I mean that this is two days later, the email has still not arrived yet.

No the mail server was not having a hiccup. Yes I did try resending it five minutes later. Yes I did check the Spam folder. No it’s not graylisting. My email address is served by G Suite, which means it’s more reliable than a normal Gmail address. Revolut can’t seem to be able to send email to Gmail. And it’s not just me. The same problem with email not arriving happened a number of months ago to my girlfriend, while sending money to my Revolut account! Anyway the answer is that I now have £50 that I can’t seem to be able to send to my sister, she ended up asking our mum for the transfer instead, and I have even less trust in the service.

I complained on Twitter about this, but without tagging in the Revolut account. When this happened to my girlfriend, and I ranted at them about it, they kept insisting to “check [my] spam folder”, which of course we did. If I asked now, I’m expecting to hear that “PSD2 made them do it”.

It’s sad, but I can’t really expect much better from a service that, despite a lot of nice ideas at the start, appear to have found a business model only to augment banks in places where high street has no offering (Ireland), or for people who can’t seem to know better (the whole Bitcoin/cryptocurrency part, that appears to be the sole attraction for Premium/Metal for quite a few people).

Opinion: FinTech vs High Street

If you’re a regular reader of this blog, you may have noticed that I have strong opinions regarding consumer financial services, particularly when it comes to Revolut, which I wrote about a lot by now.

I didn’t start writing about these services because of a professional interest, but rather because when I moved from Italy to Dublin (via Los Angeles), I felt like I stepped back ten or more years with the banking system. And while this improved significantly when I moved to London, there are still a few things baffling me from time to time.

But as I discussed in one of my recent Revolut-bashing posts, compared to Ireland the high street banking options in London are so much more interesting that I’ve effectively ditched Revolut for day-to-day payments. So why would anyone care about FinTech products?

I have been thinking this for a while, not just as a customer, but with an awareness that, if I decided to change my perspective in life and go for a riskier professional position, from my rather cushy one, FinTech appears to be the place to be right now. Particularly given the unfortunate experience I have gained in this field by now.

One of the issues appears to be one of branding, and trust. Quite a few people appear to have a dislike for high street banks because of their association with previous scandals or news. And that’s what makes it funny to see how high street banks appear to just want to enter the market with new brands.

Another thing that Monzo appears to capitalize on, in their tube advertisements, is the ability to receive instant notification of the money spent. And that’s something that I deifnitely can relate to. This is particularly important when you get to more shady stores, or to coffee stores with untrained staff, that may suggest that a transaction didn’t really go through, and suggest you to pay cash instead, charging you twice.

Indeed, this was one of the biggest advantages of using Revolut for me in Ireland. The “famous” Tesco Bank credit card didn’t really have even an online banking platform, and the only way for me to confirm whether a transaction went through was by looking at my Tesco points statements. But this is not something revolutionary: I had notifications of all online transactions, and card-present transactions over €50, on my Italian pre-paid card in 2006 (via SMS, not via app at the time, of course.)

While I feel Monzo is right to take a swing to most high street banks for not implementing these notifications, even in 2019 London it’s not true that you need to “go FinTech” to have this level of support. My American Express does the same, and you cannot say that AmEx is a new player on the market!

And it doesn’t stop at just sending me notifications for the charges: American Express goes one step further, and integrates with Google Pay so that you get the notifications even without having the American Express application installed.

Indeed, I have a feeling that, for the most part, customers would be happy if the level of support in high street banking was on par with American Express:

Their website lets you log in with a simple username/password combination, rather than the silly security theatre of “Give me the 1st, 2nd, 123th character of your password, and 1st, 5th and 6th digit of your PIN” (seriously, setting aside the random index selection, why on Earth do you need two equivalent factors?)
New charges on the card are notified immediately, either through app or through Google Pay (I don’t know about Apple Pay but I assume that’s the case there as well).
You can get your card’s PIN online, which is usually verified by a text message OTP.

One of the things that AmEx does not do, that I think all of the FinTech players appear to do, is freezing/unfreezing the card on the fly. A feature that Barclays has been advertising all over as if they had invented it.

It is pretty much possible, or certain, that some UK high street banks already started providing all of these options, maybe in different combinations. As I said, Barclays does appear to have the ability to freeze/unfreeze the card. Fineco does not mail out the PIN but rather has you requesting it online and delivers it as text message. And as I made as a point before, Santander has a credit card with no foreign transaction fees.

Many of the articles I read over the importance to FinTech startups imply that the main reason why big banks can’t be this flexible or “innovative” is that they have old, heavy and difficult to manage backends. From second hand discussions, I can believe that the backends are indeed as heavy and clunky as they are purported to be, but it does seem to me that many of the features involved can’t be that tied to the backends, given that most of the banks can provide those features already.

A number of features that I see being deployed throughout different banks is the ability to “budget” expenses. While they sound particularly interesting, this appears to be mostly a “frontend” feature. Santander has this feature, but somehow they decided to implement this on a separate Android app only, which I gave up on. Indeed, it does not allow you to correct their classification of expenses, which makes it pretty much useless, not just because some vendors are classified completely wrong, but also because sometimes the same vendor might be used for different reasons (Boots, CVS, Walgreens, and similar all provide both medicines and groceries; how you categorize their spend depends on what you bought!)

While Santander have already won me over as a bank customer, I do feel that they would win over more of my credit card expenses from American Express if they implemented “this one weird trick” of informing me of charges as they happen. Because small things like that are one of the reasons I use my AmEx quite a lot in the UK, even after I reach the needed spend to upgrade my Marriott membership to gold.

So yeah, my hope is that high street banks will finally see the competition from FinTech as a list of features that they should, opportunistically, implement, rather than an excuse for the branding and marketing departments to come up with new ideas to be “hip”.

Speaking of Foreign Transaction Fees

In the previous post about Revolut, I have left open a topic that I wanted to move to its own post: foreign transaction fees.

For those who are not acquainted with the terminology here, with foreign transaction fee I’m referring to the additional fee levied by banks and payment card companies when you incur expenses in a different currency than the one the card was issued for. Sometimes (particularly in UK and Ireland) this is referred to as an “overseas transaction fee” — which is confusing, particularly for Ireland, where the fee is applied for expenses in GBP (which is not overseas, but rather “up the road”), but not in EUR (which is mostly oversea).

This is a different cost incurred than the possible bad exchange rate that the financial institution may be applying, and it has nothing to do with the various DCC scams that you may run into when going to touristy destinations with a non-local card, although there is a link there: even online, services may suggest you to apply the charge in your local currency to avoid foreign transaction fees — as you can see in the linked post, that’s rarely a good idea, with a few exceptions (e.g. PayPal actually applies sane conversion fees in my experience, even if not the best ever).

These foreign transaction fees are set by the card issuers, and vary widely. I have seen cards with up to 6% “fex fees”, but that was back in Italy (why I say that will be clearer in a moment). In Ireland, with the exception of various fintech companies, the typical fex fees were of 2-3% — I was very happy with Tesco Banks‘s 1.75% fex fee (Tesco Bank no longer operates in Ireland.) In the UK, it appears most cards either have 0% fex fee, or 2.99% fex fee; there are a few divergences, but those two appear to be the most common options.

The reason why I am specifying this information with a country attached is that, in addition to telling you what the currency is, the mix of local-vs-foreign spend for the average person is also connected to the country. For instance, for my friends and family living in Italy, foreign transaction fees only exist when buying from foreign websites (or eBay), or when going on a “far” trip — Croatia and Switzerland being the closest countries that incur the fex fee. On the other hand, if you live in Ireland, you’ll probably have at least one recurring expense in GBP — depending on how Brexit is going to go this may change.

Indeed, for electronics you often need to look at the UK, rather than the continent — because of plugs, regulations, availability, etc. And quite a few eShops with presence both in the continent and the UK used to refuse you service from the European website, referring you to the UK one instead — this is another thing that may change after Brexit. There is a reason why, when discussing markets, most companies call it “UKI”.

I’m told that a similar situation exists for those living in Switzerland, and I can imagine this goes similar in the Nordics, given that Denmark, Sweden, and Norway have their own currencies as well, and likely a lot of services overlap.

In the UK (and again this may change after Brexit), you may very well never spend money outside of GBP because all the services exist within the country. Unless you’re an expat, in which case you’re probably still visiting the continent (Eurozone or not) fairly often, or may be paying for ongoing services (such as cellphone contracts) in that currency. This probably explains why the two sets of fex fee groups: if you’re part of the first group, you probably don’t need a card with no foreign transaction fees — while you really do in the latter case.

In my case, I have two credit cards: one from Santander, which I spoke of last time, with no foreign transaction fee, and an American Express with a 2.99% foreign transaction fee. I effectively spread the expenses on the two cards, depending on where I am — namely I try to use the Amex in the UK, and the Santander anywhere the other does not work. I could give up on the Amex, as the Santander is strictly a superset usage, but the perks provided by Amex are worth having. And that’s the most important thing: cards have perks, so you should probably consider those as well.

Thus the utility of fintech services like Revolut and Curve depend on the country you live in not just because it sets the band for foreign transaction fees, but also because they set the tone of foreign currency usage. In the UK, with the wide availability of debit and credit cards with no foreign transaction fees, their services are likely less useful than in other countries — except when it comes to perks. Indeed in the case of Curve, you would be able to keep most of the perks of a credit card, such as cashback, even if the card comes with a hefty foreign transaction fee. Except for Amex of course.

But is it convenient for you to pay for such a service? That’s another very good question. And to answer it, I’ll try to forget about the UK and go back to Ireland — mainly because here, as I now repeated a number of times, cards with no foreign transaction fee exists and you can just use one of those. Metro Bank has free current accounts with cards that come with cards without foreign transaction fees in Europe. Santander has a £3/month credit card with no foreign transaction fees, and 0.5% cashback. Halifax has a Clarity MasterCard that comes with no monthly fee, no foreign transaction fees (and of course no perks.)

But let’s go back to Ireland and take a look at the options. As I said the usual foreign transaction fee in the country was between 2% and 3%. In the case of Ulster Bank, the card I used to have had 2.75% foreign transaction fee. At which point would it have been cheaper for me to subscribe to Curve Black, at €9.99/month, rather than give Ulster Bank their fees? (And for simplicity here, I’m not talking about exchange rates; the exchange rate for their MasterCard is network-provided so it’s not at all bad, and in fact it’s comparable to Revolut’s.)

As most services would require a yearly commitment, we should consider the spend on an yearly basis too. This makes the cost €119.88, but we’ll call it €120 to make it easier to run umbers on them. Let’s just call the twelve cents a rounding error. If we’re ignoring the cashback options (as in Ireland there were none, beside Tesco Bank), the amount of foreign expenses you’d need to break even on Curve black with the foreign transaction fee noted above is about €4364 (divide the yearly cost by the foreign transaction fee). That’s the cost of fairly big vacation for a family (note that you can’t include flights in the vacation cost, as those would be billed by the currency of the country of origin, which is likely local).

If you have a card that provides cashback, then things become more complicated, because you’d have to include the cashback in the calculation. If you’re curious the following formula will give you the number, making S the yearly subscription cost of the service, F the foreign transaction fee percentage, and C the cashback percentage:

(S + (S/F) * C) / F

For Revolut Metal, with their variable cashback, figuring out the number is a bit more annoying. But we’re also talking about 1% in the best case scenario (all non-European spend). So the basic number (€5673) only goes down to €5616. The 0.1% cashback option of all European spend is so minimal that it’s not worth calculating exactly.

So what should you do if you don’t usually spend that kind of money on foreign transactions? You can still use the Revolut and Curve and other fintech services without paying for them, and grab the best deal you can until they go bust. Or if you don’t want to bother, you can just spend on your normal cards, get your usual perks and ignore the need for no foreign transaction fees.

Indeed, if your options are spending on Curve attached to a debit card with no cashback and no perks, or spend on an American Express Platinum Cashback Credit Card, you would need to spend more than £5330 a year in foreign transactions for it to be worth it — and that’s assuming you don’t qualify for the higher tier. And this is probably the worst case scenario for the UK, for a non-zero foreign transaction fee card.

Is Revolut Still a Good Thing?

You may remember that a few years ago I wrote a positive review of Revolut, the fintech startup that provides payment cards with stored value and no foreign transaction fees. I have been using it for a long time by now, and had mostly stood by that review, until the second half of last year, where things started to appear more complicated. Given the current flurry of stories on the company, from silly advertising shenanigans to uncovering of poisonous working conditions, I thought it would be a good time to write some more up to date words, as I don’t think I can recommend Revolut as much as I did before anymore.

First of all, I started feeling uneasy recommending Revolut since they started down the path of selling cryptocurrencies as an added-value feature. I hold a personal belief that participating in the trading of Bitcoin and other similar “currencies” is unethical (see Thomas’s rant on the topic), and I don’t like being associated with companies focusing on them. I have looked the other way for a while, though, because I knew that using the words “cryptocurrency” and “blockchain” make money appear out of nowhere for most startups, even when there’s no rhyme or reason for it. I just had a bad taste in my mouth for this.

The problem is that Revolut, even when I had the Premium version, built something very cool, but a bit rough around the edges. And as a customer, it is annoying to see them jumping the shark onto cryptocurrencies, instead of making location-based security actually reliable, implementing 3DSecure/VBV integrations, or finding a way to get a proper banking license and FSCS insurance (all of which would be requirements for me and most people to use Revolut as a replacement for high-street banking).

Instead, what we see is that Revolut is adding “features” trying to upsell you into their premium services. This is not entirely bad, because you need paying customers to run a business. Unfortunately my impression is that they offered and offer so much on their free tier, that they are tackling on random stuff that has nothing to do with banking itself, just to get people to sign up for their Premium and Metal tiers.

As an aside, I still don’t understand this trend of providing heavy (“18g” as they boast some companies) metal cards. The last thing I want from a credit card is to be heavy, as I barely even want to have to take it out. I’m all in favour of the trend of not embossing the name and number, preferring to print it on the back, but it does not need to be metal for it. Indeed, Curve (that I’ll get again in a moment) did exactly that.

We’ve just come back from a trip to the Continent, and what we did notice that Revolut tried to upsell us medical and travel insurance at every change of country (even when we just connected flights through third countries). This is not just annoying as we’re not interested in it (we’re European citizens, visiting European countries, and work provides both of us with a basic travel insurance), but it’s also annoying because it makes use of the location information, which I provide for the security feature, for marketing. Similarly, I recently had more notifications about them trying to upsell me Metal than actual transactions.

For a while, I actually did pay for the Premium service. Mostly under the idea of “putting my money where my mouth is”, that is to make sure that the company could keep operating a service I loved. Unfortunately it turned out a bad idea: not just because Revolut cannot replace a high street bank in the UK (no FSCS to protect your account, no BACS direct debits, etc), but also because the Premium “perks” were not something I cared about, and the dedicated service team was still useless when it came to even telling me the top-up limits when I changed the card I used for top-up.

If you already have two physical cards (and paid for it), you need to pay to replace one of them with a Premium card, if you so wish (but it gains nothing but a different colour, so I never did that). The unlimited exchange is not particularly useful when you already don’t reach the free tier’s spend, and the ATM limits is only useful if you plan to actually use cash, which I really try not to. The one interesting feature that is advertised for Premium customers, but as far as I can tell is also present as a one-off charge for non-Premium one, is the disposable virtual card, that changes PAN every time you use it. But even that is not as secure as it looks, as I’m told that vendors are still able to charge again a disposable card that already changed number.

Okay admittedly there’s the travel and medical insurance, but as I said earlier, I get a better travel medical insurance from work (and probably there’ s better out there) and a credit card such as American Express would provide a better baggage/flight insurance. This is very subjective of course, it’s well possible that for other people, with other employers, and in other countries, these insurances are actually worth it.

Speaking of circumstances, I think I might not have felt so strongly against Revolut if I was still in Ireland. Not just because they seem to have implemented SEPA DD Core support, so you can actually use it to pay your bills there, but also because the alternatives of high street banking there are significantly worse than here.

In London, I now settled on Santander as my primary bank, both for the current account and for a 0% foreign transaction fee credit card, their All-in-One Credit Card. These come to £5 per month for the account, and another £3 per month for the credit card (compare against Revolut’s premium at £6.99 and Metal tier at £12.99), and while the free foreign ATMs withdrawal are limited to Santander’s own network (limiting the countries you can use them on), this is a full-featured, FSCS-insured account, with cashback, retailer offers, and active interest on the current account’s deposit. If you don’t want (or can’t afford) a credit card, Metro Bank offers 0% foreign transaction fee for European transactions on their free accounts’ debit cards. And I’m sure that other banks have similar arrangements all over the place. Basically, the UK has a significantly wider range of offers, that make Revolut less necessary than in Ireland.

But even for Ireland, and for other countries that do not have such a selection of high-street banks, Curve – that I complained about before – decided to change their target marketing a bit, now offering a “front” for any Visa and MasterCard card to provide 0% foreign transaction fee, with their premium option existing to raise the limit of monthly transactions. That would have been something awesome to have when I lived in Dublin, to keep getting Tesco points, while not paying the 1.75% of foreign transaction fee on their credit card. (If you are interested to try that, my referral code is BG2G3).

Both Curve and Revolut have a Metal card with which they provide cashback. In the case of the former, these are retailers-limited, and I can only assume they are based on some third party’s selection of perks, as the retailers are pretty much the same that Santander and Lloyd’s provide retailers offers for. Revolut instead provides cashback on all spend, 0.1% on European spend, and 1% for non-European spend (although there does not seem to be an obvious definition of Europe on their marketing material, I assume it’s deep into the terms of service).

While cashback is always a nice bonus, it only makes sense if you can break even on the cost of one’s service by spending. With Revolut Metal, that would be an astounding £13k (thirteen thousands pound) per month in European spend, or £1299 of non-European spend. I do know some extremely frequent travellers to the States or Asia that would be able to spend the latter, but that’s more of an exception than a rule. And if you can spend the former, you probably can get more than that in interest by keeping the money in an active-interest current account, and paying with a normal credit card.

For comparison, Santander’s card I linked above costs £3/month (you don’t even need their bank account). It has 0% foreign transaction fee on all spend. And a cashback of 0.5% (five times Revolut’s European cashback) on all spend. It takes only £600 a month to break even, and that’s without counting additional retailer offers, or additional perks from their current accounts.

And even if you look at American Express (which is never considered a cheap option) and their cashback options, the numbers are significantly different. Their Platinum cashsback card is £25 per year, and includes a better travel insurance, 1% cashback on all spend to £10k and 1.25% over that. Plus retailers offers and supplementary cards for the family. Although be warned if you want to go down that road, that American Express charges you 2.99% foreign transaction fees, for every single one of their cards in the UK.

I was going to take a detour talking about foreign transaction fees, but I will leave it for another post, because it’s a lot of content, and a lot of explanation to be done there.

So the final words of this post are: I’m not sure I trust Revolut anymore. They seem to be taking “marketing risks” to get people to pay for services, but at the same time there’s very little value in their paid services. I don’t think that the company will be able to sustain the current trajectory without venture capital money, and I find scary the idea of relying on a VC-funded pseudo-bank for my own money.

Update (2019-03-27): just a few days after I wrote this blog post, I received two email from Revolut, with widely different content, that I think merit a bit of description, thus why this update.

The latest email is an announcement of new details (new sort code and account number) for their GBP accounts. This is effectively a change in intermediary bank that maintains the GBP account proxies for Revolut. Nothing particularly eventful in by itself, but there are a few notable things. The announcement is declared “great news” for their customers, but it also highlight yet another feature that high street banking would have, and Revolut lacks: redirections.

When you switch bank account with a high street bank, the bank will take care of moving standing orders, direct debits, automatic salary payments, and redirect any transfer to the old bank account to the new one. Revolut is instead telling all the customers that they have to deal with all the required changes of both payment and transfer. Not just that, but they don’t appear to guarantee any specific grace period in which both accounts would exist: they say that the new details will appear in the app before May 22nd, which is when the old account will stop working:

⚠️ Your old account details will stop working from the 22nd May 2019.

Salaries and standing orders

If you receive your salary into your Revolut account, you’ll need to send your new account details to your employer before the 22nd May. Again, we’ll let you know as soon as they arrive.

For standing orders from your external bank to your Revolut account, you’ll need to update your bank with your new details before 22nd May. For recurring payments set up from your Revolut account to another bank, you don’t need to do anything.
Revolut email arrived on 2019-03-27

To give you an idea of time frame involved, the company I work for freezes the salary payment details around the 5th of the month for payments on the 25th. This means that if the new details arrive after 5th of May, and you’re paid monthly, you may be unable to receive the salary. Hopefully, the old accounts would just reject the transfer, but even in that case, retrieving the missing salary can easily take two weeks, which for a number of people would be a significant risk.

For comparison, the previous email I received just twenty hours before, also from Revolut, had as subject «👕Should we release Revolut merch?». This is a company that just before announcing a significant disruption of service, that a high street bank would never subject their customers to, asks whether you would like to wear their brand around, making yourself not just a product, but a walking billboard.

Update 2019-01-04: see also the October update.

A review of the Curve debit card

Somehow, I end up spending a significant amount of my time thinking, testing and playing with financial services, both old school banks and fintech startups.

One of the most recent ones I have been playing with is Curve. The premise of the service is to allow you to use a single card for all transactions, having the ability to then charge any card underneath it as it is convenient. This was a curious enough idea, so I asked the friend who was telling me about it to give me his referral code to sign up. If you want to sign up, my code is BG2G3.

Signing up and getting the card is quite easy, even though they have (or had when I signed up) a “waitlist” — so after you sign up it takes a few days for you to be able to actually order the card and get it in your hands. They suggest you to make other people sign up for it as well to lower the time in the waitlist, but that didn’t seem to be a requirement for me. The card arrived, to my recalling, no more than two days after they said they shipped it, probably because it was coming from London itself, and that’s all you need to receive.

So how does this all work? You need to connect your existing cards to Curve, and verify them to be able to charge them — verification can be either through a 3Dsecure/Verified by Visa login, or through the usual charge-code-reverse dance that Google, PayPal and the others all use. Once you connect the cards, and select the currently-charged card, you can start using the Curve card to pay, and it acts as a “proxy” for the other card, charging it for the same amount, with some caveats.

The main advantage that my friend suggested for this set up is that you if you have a corporate card (I do), you can add that one to Curve too, and rely on that to not have to do the payback process at work if you make a mistake paying for something. As this happened to me a few times before, mostly out of selecting the wrong payment profile in apps such as Uber or Hailo, or going over the daily allowance for meals as I changed plans, it sounded interesting enough. This can work both by making sure to select the corporate card before making the purchase (for instance by defaulting to it during a work trip), or by “turning back time” on an already charged transaction. Cool.

I also had a hope that the card could be attached to Google Pay, but that’s not the case. Nor they implement their own NFC payment application, which is a bit disappointing.

Beside the “turn back time” feature, the app also has some additional features, such as the integration with the accounting software Xero, including the ability to attach a receipt image to the expense (if this was present for Concur, I’d be a real believer, but until then it’s not really that useful to me), and to receive email “receipts” (more like credit card slips) for purchases made to a certain card (not sure why that’s not just a global, but meh).

Not much else is available in the app to make it particularly useful or interesting to me, honestly. There’s some category system for expenses, very similar to the one for Revolut, but that’s about it.

On the more practical side of things, Curve does not apply any surcharges as long as the transaction is in the same currency as the card, and that includes the transactions in which you turned back time. Which is handy if you don’t know what the currency you’ll be charged in will be in, though that does not really happen often.

What I found particularly useful for this is that the card itself look like a proper “British” card — with my apartment as the verified address on it. But then I can charge one of my cards in Euro, US Dollars, or Revolut itself… although I could just charge Revolut in those cases. The main difference between the two approach is that I can put the Euro purchases onto an Euro credit card, instead of a debit one… except that the only Euro credit card I’m left also has my apartment as its verifiable billing address, so… I’d say I’m not the target audience for this feature.

For “foreign transactions” (that is, where the charged currency and the card currency disagree), Curve charges a 1% foreign transaction fee. This is pointless for me thanks to Revolut, but it still is convenient if you only have accounts with traditional banks, particularly in the UK where most banks apply a 3% foreign transaction fee instead.

In addition to the free card, they also offer a £50 (a year, I guess — it’s not clear!) “black” card that offers 1% cashback at selected retailers. You can actually get 90 days cashback for three retailers of your choice on the free card as well, but to be honest, American Express is much more widely accepted in chains, and its rewards are better. I ended up choosing to do the cashback with TfL, Costa (because they don’t take Amex contactless anyway), and Sainsbury’s just to select something I use.

In all of this, I have to say I fail to see where the business makes money. Okay, financial services are not my area of expertise, but if you’re just proxying payments, without even taking deposits (the way Revolut does), and not charging additional foreign transaction fees, and even giving cashback… where’s the profit?

I guess there is some money to be made by profiling users and selling the data to advertisers — but between GDPR and the fact that most consumers don’t like the idea of being made into products with no “kick back”. I guess if it was me I would be charging 1% on the “turn back time” feature, but that might make moot the whole point of the service. I don’t know.

At the end of the day, I’m also not sure how useful this card is going to be for me, on the day to day. The ability to have a single entry in those systems that are used “promiscuously” for business and personal usage sounds good, but even in that case, it means ignoring the advantages of having a credit card, and particularly a rewards card like my Amex. So, yeah, don’t really see much use for it myself.

It also complicates things when it comes to risk engines for fraud protection: your actual bank will see all the transactions as coming from a single vendor, with the minimum amount of information attached to it. This will likely defeat all the fraud checks by the bank, and will likely rely on Curve’s implementation of fraud checks — which I have no idea how they work, since I have not yet managed to catch them.

Also as far as I could tell, Curve (like Revolut) does not implement 3DSecure (the “second factor” authentication used by a number of merchants to validate e-commerce transactions), making it less secure than any of the other cards I have — a cloned/stolen card can only be disabled after the fact, and replaced. Revolut at least allows me to separate the physical card from my e-commerce transactions, which is neat (and now even supports one time credit card numbers).

There is also another thing that is worth considering, that shows the different point of views (and threat models) of the two services: Curve becomes your single card (single point of failure, too) for all your activities: it makes your life easy by making sure you only ever need to use one card, even if you have multiple bank accounts in multiple countries, and you can switch between them at the tap of a finger. Revolut on the other hand allows you to give each merchant its own credit card number (Premium accounts get unlimited virtual cards) — or even have “burner” cards that change numbers after use.

All in all, I guess it depends what you want to achieve. Between the two, I think I’ll vastly stick to Revolut, and my usage of Curve will taper off once the 90 days cashback offer is completed — although it’s still nice to have for a few websites that gave me trouble with Revolut, as long as I’m charging my Revolut account again, and the charge is in Sterling.

If you do want to sign up, feel free to use BG2G3 as the referral code, it would give a £5 credit for both you and me, under circumstances that are not quite clear to me, but who knows.