The Price of Moving Money
In January 2023, a quiet milestone passed with almost no fanfare: for the first time, the average fee Wise charged to move money across borders fell below 0.62% of the transfer value. Not 6.2%. Not 2%. Not even the 1% that most fintech competitors advertised as revolutionary. Six-tenths of a cent on every dollar — a number so low it would have been dismissed as operationally impossible a decade earlier, when the global average cost of sending $200 internationally hovered around 9.3% and Western Union's shareholders considered that margin a birthright. The number kept falling. By mid-2024, Wise's take rate had compressed further to roughly 0.57%, a figure that approaches the theoretical floor of what it costs to actually execute a cross-border payment when you strip away every intermediary, every correspondent banking chain, every hidden markup buried in the exchange rate. The company was not losing money. It was, in fact, generating over £1.1 billion in revenue and posting comfortable operating margins north of 20%. This is the central paradox of the Wise story: a company that has spent its entire existence making itself cheaper, that has turned relentless price reduction into a growth engine, and that has somehow built a durable business — a genuinely profitable, publicly traded, increasingly infrastructure-like business — on the premise that the product should cost almost nothing.
The conventional wisdom in financial services is that you own the spread. You capture the delta between what it costs you to move, store, or lend money and what you charge the customer. The wider the spread, the better the business. Wise inverted this logic so completely that its entire competitive position depends on the spread being as narrow as possible — because the narrower the spread, the more volume flows through the system, the more the infrastructure amortizes, the more the network effects compound, and the harder it becomes for anyone else to match the price and the experience simultaneously. It is a flywheel built on absence — on the systematic elimination of the very fees that fund the traditional financial system.
By the Numbers
Wise at Scale
£1.13BRevenue, FY2024 (year ending March)
12.8MActive customers
£118.5BTotal volume processed, FY2024
0.57%Average take rate (and falling)
~22%Adjusted operating margin, FY2024
~6,000Employees worldwide
£6.5BApproximate market cap (LSE, mid-2024)
65%Transfers arriving instantly or within an hour
Two Estonians and a Spreadsheet
The founding story has been told often enough to acquire the texture of myth, but the underlying economics remain instructive. In January 2011, two Estonians living in London — Taavet Hinrikus and Kristo Käärmann — discovered they shared a problem that neither the world's largest banks nor its most sophisticated payment networks had bothered to solve elegantly.
Hinrikus had been Skype's first employee, employee number one at a company that would sell to eBay for $2.6 billion and then to Microsoft for $8.5 billion. He earned his salary in euros but lived in London, paying rent in pounds. Every month, his bank converted his euros at a rate that looked reasonable until you compared it to the mid-market rate published by Reuters — at which point you discovered the bank was silently extracting 3–5% through exchange rate markup, on top of whatever explicit fee it charged. Käärmann, a management consultant and former Deloitte and PwC alumnus, had the inverse problem: paid in pounds, with a mortgage in Estonia denominated in euros. Same hidden tax, different direction.
The solution they devised was disarmingly simple. Käärmann would deposit pounds into Hinrikus's UK bank account at the real mid-market exchange rate. Hinrikus would deposit the equivalent euros into Käärmann's Estonian account. No money actually crossed a border. Both got the rate they deserved. The banks' margin evaporated.
They called the resulting company TransferWise — the name itself a small act of provocation, implying that the alternative was TransferStupid — and launched it in early 2011 with a peer-to-peer matching model that replicated their kitchen-table solution at scale. The insight was not technological but structural: the vast majority of cross-border payment fees were not operational costs but rents extracted by institutions exploiting information asymmetry and customer inertia. The mid-market rate was public. The bank's markup was hidden. TransferWise would make it visible.
Banks have gotten away with this for decades because most people don't even know they're being charged. The fee is invisible — it's hidden inside the exchange rate. Our job is to make it visible, then eliminate it.
The Peer-to-Peer Illusion and the Infrastructure Reality
There is a persistent misconception about Wise's early model that needs correcting, because it reveals something important about how the company actually works.
The peer-to-peer matching narrative — your pounds stay in the UK, someone else's euros stay in Europe, the system just nets the flows — is elegant and was genuinely how some early transfers operated. But it was never the whole story, and it was never the business's limiting factor. In practice, perfectly matching bidirectional flows in real time across dozens of currency corridors is extraordinarily difficult. The GBP-to-EUR corridor might balance nicely; the GBP-to-INR corridor almost certainly does not, because the flow is overwhelmingly one-directional (diaspora remittances). What Wise actually built, progressively and with increasing sophistication, was a network of local bank accounts in dozens of countries, pre-funded with local currency, that allowed it to settle transfers domestically on both ends — your pounds go into Wise's UK account, and Wise's Indian account sends rupees to the recipient. The cross-border leg happens in bulk, at wholesale rates, when Wise rebalances its liquidity pools.
This is a fundamentally different business than peer-to-peer matching. It is a treasury management operation layered on top of a consumer-facing transfer product. It requires banking licenses or partnerships in every corridor. It requires maintaining liquidity buffers in dozens of currencies simultaneously. It requires real-time FX risk management. And it creates a structural cost advantage that compounds with scale — because the more volume flows through the network, the more the liquidity pools naturally balance, the less Wise needs to access wholesale FX markets, and the lower the marginal cost of each additional transfer.
By 2015, the company had built local payment infrastructure in more than 50 countries. The matching narrative persisted in marketing materials and press coverage long after the actual settlement architecture had evolved beyond it. This is not unusual — companies often grow past their origin stories before anyone notices — but in Wise's case, the gap between narrative and reality mattered because the real model was far more defensible than the story suggested. Peer-to-peer matching is easy to replicate. A globally distributed, locally settled, algorithmically rebalanced liquidity network is not.
The Provocateur's Playbook
The early years of TransferWise were defined as much by marketing audacity as by product quality. This was deliberate — and it reflected Käärmann's particular genius for translating a complex financial argument into visceral, shareable outrage.
In 2014, TransferWise hired a team of actors to strip to their underwear outside the Royal Exchange in London, holding signs that read "Nothing to hide" — a pointed reference to the hidden fees banks buried in their exchange rates. The stunt earned global media coverage worth many multiples of its cost. In 2015, they ran full-page newspaper ads headlined "Hey banks, the game is up" with detailed comparisons showing how much major UK banks charged versus the mid-market rate. Barclays, HSBC, Lloyds — all named, all shamed. The banks, predictably, did not respond. Their silence was the response; they could not defend the spread.
This confrontational posture served multiple strategic purposes simultaneously. It generated awareness at a fraction of traditional customer acquisition cost. It framed the competitive dynamic not as "fintech vs. fintech" but as "transparency vs. the banking cartel" — a framing that made TransferWise the protagonist of a populist narrative that was particularly resonant in post-financial-crisis Britain. And it established a brand identity rooted in missionary zeal rather than financial product marketing, which attracted a specific type of customer: the convert who would evangelize the service to friends and family, creating organic referral loops that became the company's most efficient growth channel.
Richard Branson invested in the company's early rounds — the alignment between Branson's antiestablishment brand and TransferWise's insurgent positioning was not accidental. Andreessen Horowitz led the Series C in 2015 at a $1 billion valuation, marking TransferWise as the first European fintech "unicorn" focused on payments.
Peter Thiel's Valar Ventures had been in since the Series A. The investor roster read like a deliberate composition: Silicon Valley's most contrarian capital backing Europe's most confrontational fintech.
I've seen first hand how much traditional banks charge to move money across borders. TransferWise is cutting through all of that.
The Margin Compression Engine
Most companies talk about lowering prices. Wise actually does it — systematically, publicly, and as a matter of articulated corporate philosophy. This is not a marketing position. It is the company's core strategic mechanism, and understanding it requires understanding the relationship between price, volume, and infrastructure cost in Wise's model.
Here is the logic: Wise's infrastructure costs are largely fixed or semi-fixed — the banking licenses, the local payment rails, the compliance systems, the technology platform. The marginal cost of processing an additional transfer through an existing corridor is very low. As volume grows, the average cost per transfer falls. Wise passes a portion of that cost reduction to customers as lower fees, which attracts more volume, which further reduces average cost, which enables further price reduction. It is a deflationary spiral that Wise has engineered to run continuously.
The data tells the story with precision. In FY2017, Wise's average take rate was approximately 0.74%. By FY2020, it had fallen to 0.68%. By FY2022, 0.63%. By FY2024, roughly 0.57%. Each year, Wise processes more volume at lower unit economics, yet total revenue grows because volume growth outpaces price compression. In FY2024, total cross-border volume reached £118.5 billion — a 24% increase year-over-year — while revenue grew 24% to £1.13 billion despite the continued decline in take rate. The company is getting cheaper and bigger simultaneously.
📉
The Take Rate Trajectory
Wise's average fee as a percentage of transfer value
| Fiscal Year | Avg. Take Rate | Total Volume | Revenue |
|---|
| FY2017 | ~0.74% | ~£12B | ~£100M |
| FY2019 | ~0.71% | ~£34B | ~£240M |
| FY2021 | ~0.65% | ~£54B | ~£421M |
| FY2023 | ~0.62% | ~£96B | ~£847M |
| FY2024 | ~0.57% |
This pattern — volume growth systematically outpacing price reduction — is the financial signature of a company approaching infrastructure status. It mirrors the dynamic that Amazon Web Services exhibited in its scaling years: relentless price cuts that expanded the addressable market faster than they compressed margins. The crucial question for Wise is whether the take rate has a floor, and what happens to volume growth as the company exhausts easy corridors and begins competing for increasingly price-sensitive segments of the market. More on that later.
The Platform Turn
For its first several years, Wise was a single-product company: you sent money abroad, Wise moved it, you paid a fee. Clean. Simple. Dangerously dependent on a single revenue stream.
The pivot — or more accurately, the expansion — began in earnest around 2017 with the launch of the Wise (then TransferWise) multi-currency account and debit card. The logic was straightforward: if Wise had already built the infrastructure to hold and move dozens of currencies, why not let customers hold those currencies directly, spend them abroad at the mid-market rate, and receive money from foreign sources into local bank details? The borderless account, as it was initially branded, transformed Wise from a transfer service into a multi-currency banking layer — a single account with local bank details in major markets (GBP sort code and account number, EUR IBAN, USD routing and account number, AUD BSB) that could hold, convert, send, and receive in 40+ currencies.
This was a category-defining move. Suddenly, Wise was not just competing with Western Union and bank wire transfers. It was competing with Revolut, N26, and the neobanking wave for the primary financial relationship of internationally mobile consumers, freelancers, and small businesses. The debit card — Visa-branded, with conversion at the mid-market rate plus a small fee — gave customers a reason to keep money in the Wise ecosystem rather than converting and withdrawing.
The strategic implications were profound. An account is stickier than a transfer. A customer who holds balances, receives salary, and spends on a Wise card is orders of magnitude harder to churn than one who uses Wise twice a year to send money to family. Customer balances grew from negligible amounts in 2018 to over £12 billion by early 2024 — money sitting in Wise accounts, generating interest income that the company shared with customers (a Wise innovation that predated the broader neobank trend of interest-bearing current accounts) while also funding the liquidity pools that made the core transfer product cheaper. The flywheel again: more customers → more balances → cheaper funding → lower transfer fees → more customers.
Wise Platform: Selling Picks During the Gold Rush
If the multi-currency account was the first expansion, Wise Platform was the second — and arguably the more consequential for long-term competitive position.
Launched in 2019 and scaled aggressively through 2021–2024, Wise Platform (originally called TransferWise for Banks) offered Wise's cross-border payment infrastructure as an API that other financial institutions could embed directly into their own products. The pitch to banks was uncomfortably direct: your cross-border payment product is terrible and expensive; embed ours, brand it as yours, and your customers get instant, cheap transfers while you stop losing them to us. The pitch to fintechs and neobanks was gentler: building cross-border payment infrastructure from scratch takes years and hundreds of millions; use ours instead and focus on your differentiation.
The client list grew to include some genuinely significant names: Monzo and N26 among the neobanks, Google Pay for international remittances, and dozens of smaller banks and financial institutions across multiple continents. By FY2024, Wise Platform was processing meaningful volume — the company reported that platform partnerships contributed to its growth though exact revenue breakdowns remained somewhat opaque.
The strategic genius of Wise Platform is that it converts would-be competitors into distribution partners. Every bank that embeds Wise's infrastructure is a bank that will not build its own — and that sends volume through Wise's network, improving the liquidity pools and cost economics for everyone, including Wise's direct consumers. It is the Amazon Web Services playbook applied to payments: productize your internal infrastructure, sell it to the market, and let your customers' volume subsidize your own cost reductions. The infrastructure becomes more defensible with every integration, because switching costs for an embedded API are dramatically higher than switching costs for a consumer transfer app.
Wise Platform now serves banks, large enterprises, and other fintechs, offering them access to our infrastructure to power fast, cheap, and transparent cross-border payments within their own products.
The Listing That Wasn't an IPO
On July 7, 2021, Wise became the largest-ever technology company to list on the London Stock Exchange — but it did so through a direct listing rather than a traditional IPO, a choice that was characteristic in both its substance and its symbolism.
A direct listing meant no new shares were issued, no investment banks earned underwriting fees, no roadshow inflated the price through artificial scarcity, and no lockup period constrained existing shareholders. It was, in essence, the Wise ethos applied to capital markets: eliminate the intermediary, make the process transparent, let the market discover the price. The company opened trading at £8.00 per share, implying a valuation of approximately £8 billion ($11 billion) — a remarkable outcome for a company that had raised a total of approximately $1.7 billion in venture funding over the preceding decade.
The choice of London over New York was itself a statement. Wise was an Estonian-founded, London-headquartered company with a global customer base, and listing on the LSE was partly about identity and partly about proving that a European tech company did not need to genuflect before NASDAQ to access public market capital. The company simultaneously obtained authorization to list American Depositary Receipts (ADRs) in the US, hedging the liquidity question without abandoning the London principle.
The early public market performance was rocky. After peaking above £11 per share in September 2021, the stock declined through the broader tech correction of 2022, hitting a low near £4 in late 2022 — a drop of over 60% from its peak. This was partly market conditions (growth multiples compressed across the board), partly Wise-specific: the company disclosed in early 2023 that Käärmann had received a personal tax penalty from HMRC for late filing, a disclosure that created governance concerns that a founder-led, dual-class share structure amplified. The stock has since recovered significantly, trading in the range of £7–9 through mid-2024, as the market absorbed the governance noise and refocused on the underlying business metrics — which, throughout the downturn, continued to improve.
The Käärmann Question
Kristo Käärmann is an unusual figure in European fintech — more abrasive than polished, more engineer than salesman, with a tendency toward blunt public statements that would make a communications team flinch. He holds a degree in mathematics and computer science from the University of Tallinn and an MBA from INSEAD, but his operating style owes more to the former than the latter. Colleagues describe a leader obsessed with cost efficiency to a degree that borders on the pathological — personally reviewing expense reports, questioning headcount additions, insisting that the company's internal cost culture mirror the frugality it promises customers.
The HMRC tax matter — Käärmann was fined for filing his personal tax return late, a disclosure the company made in January 2023 — was, by the standards of corporate governance crises, relatively minor in substance but outsized in symbolic weight. Here was the co-founder of a company built on transparency and financial responsibility, personally failing at basic tax compliance. Wise's board investigated, concluded the matter was personal rather than indicative of systemic governance failure, and Käärmann retained his position as CEO. But the episode exposed the tension inherent in founder-led companies: the same intensity and iconoclasm that drives unconventional strategic thinking can manifest as cavalier disregard for administrative norms.
Hinrikus, the other co-founder, had stepped back from day-to-day operations years earlier, though he remained on the board and retained significant shareholding through his investment firm. The company's leadership increasingly reflected Käärmann's singular vision: a missionary belief that cross-border payments should be fast, cheap, and transparent, and that Wise's job was to drive costs toward zero while building a business model that thrived at that asymptote.
The dual-class share structure — Käärmann and Hinrikus hold shares with enhanced voting rights — means that this vision is essentially unchallenged at the governance level. For believers, this is a feature: it insulates long-term thinking from quarterly earnings pressure. For skeptics, it is a vulnerability: it concentrates power in a founder who has demonstrated at least one significant lapse in personal financial diligence.
The Geography of [Friction](/mental-models/friction)
To understand Wise's competitive position, you must understand the geography of cross-border payments — a landscape shaped not by technology but by regulation, banking infrastructure, and the deeply uneven distribution of financial system modernity across the globe.
Moving money from the UK to Europe is, in 2024, fast and relatively cheap even through traditional channels — SEPA (the Single Euro Payments Area) has created near-instant, low-cost interbank transfers across the eurozone, and UK Faster Payments provides similar infrastructure domestically. Wise's advantage in these corridors is real but modest: better exchange rates, a cleaner user experience, marginally lower fees. The corridors where Wise's value proposition is most dramatic — and where the company still generates outsized margins — are the high-friction routes: UK to India, US to Philippines, Europe to Nigeria, any developed-market-to-emerging-market corridor where correspondent banking chains are long, settlement times are measured in days, and the incumbent fee structure extracts 5–10% of the transfer value.
These are also the corridors where regulation is most complex. India's Reserve Bank has historically maintained tight controls on inbound remittances and the entities authorized to process them. Nigeria's Central Bank has periodically restricted access to official exchange rates. The Philippines requires specific licensing for remittance operators. In each case, Wise has pursued direct licensing — a strategy that is slower and more expensive than partnering with local intermediaries but that yields better unit economics and more control over the customer experience in the long run.
By 2024, Wise held direct licenses or authorizations in over 30 jurisdictions. The company could send money to recipients in more than 160 countries, from accounts in more than 40 currencies. This regulatory footprint — built over more than a decade of applications, audits, compliance buildouts, and occasionally frustrating interactions with central banks — is itself a significant moat. It cannot be replicated quickly. And it creates a peculiar dynamic in which the countries where Wise's value add is highest are also the countries where the regulatory barriers to entry are steepest.
The Business Customer Nobody Expected
The initial TransferWise customer was a consumer: an expat sending money home, a freelancer receiving payment in a foreign currency, a parent paying a child's overseas tuition. The product was designed for individuals, priced for individuals, marketed to individuals.
But businesses started showing up. Small ones at first — Etsy sellers receiving payments in multiple currencies, consultants invoicing in dollars and spending in pounds, small e-commerce merchants managing multi-currency cash flows. Then larger ones. By 2020, Wise had built a dedicated business product with batch payment capabilities, multi-user access, API integrations with accounting software, and pricing tiers for higher volumes. By FY2024, Wise Business accounted for a substantial and growing share of total volume — the company reported that business customers were its fastest-growing segment, with volume growth outpacing personal transfers.
This matters enormously for Wise's long-term economics. Business customers tend to have higher average transfer values, more frequent transactions, and greater platform stickiness — a company that has integrated Wise into its payroll processing or supplier payment workflow is far less likely to switch than an individual who uses Wise once a quarter. Business customers also generate data that improves Wise's fraud detection, compliance scoring, and corridor optimization. The Wise Business product, along with Wise Platform's B2B infrastructure play, represents the company's clearest path to the kind of enterprise-grade stickiness that separates transient fintech success from durable infrastructure.
The competitive dynamics are different in business payments. Here, Wise competes less with Western Union and more with the likes of Payoneer, Airwallex, and the traditional banking treasury services of HSBC and Citi. The product requirements are more demanding — businesses need reconciliation, audit trails, multi-approval workflows, ERP integration. But the core value proposition is identical: the mid-market exchange rate, transparent fees, and fast settlement. A CFO managing $10 million in annual cross-border supplier payments cares just as much about hidden FX markups as an expat sending £500 to Tallinn. Probably more.
The [Speed](/mental-models/speed) Imperative
Price is Wise's most visible competitive weapon, but speed may be its most strategically important. In a world where domestic payments are increasingly instant — the UK's Faster Payments, India's UPI, Brazil's Pix, the eurozone's SEPA Instant — the multi-day settlement time of traditional cross-border transfers feels increasingly anachronistic. Wise has invested heavily in reducing transfer times, and the numbers reflect genuine progress: by FY2024, approximately 65% of transfers arrived instantly or within one hour, up from roughly 50% two years earlier. The target, stated publicly, is to make the vast majority of transfers arrive in under 20 seconds.
Achieving this requires not just fast payment rails on both ends but predictive compliance screening — running anti-money-laundering and sanctions checks before the customer even confirms the transfer, using machine learning models trained on Wise's enormous transaction dataset to pre-approve low-risk transfers and flag high-risk ones for manual review without slowing the entire system. It requires pre-funded liquidity pools in destination currencies so that the outbound leg can execute immediately rather than waiting for Wise to settle its own interbank position. And it requires direct integration with local instant payment networks — not going through intermediary banks that add hours or days to the settlement chain.
Speed and price are linked in Wise's model through a mechanism that is not immediately obvious: faster transfers require larger pre-funded liquidity pools, which require more capital tied up in foreign currency positions, which creates FX exposure that must be hedged, which costs money. But faster transfers also mean higher customer satisfaction, lower support costs (customers don't email asking where their money is), and higher conversion rates (a customer shown "arrives in 20 seconds" is more likely to complete the transfer than one shown "arrives in 1–2 business days"). Wise has concluded that the net effect is positive — that investing in speed is a self-funding improvement — and the data appears to support this.
We want the experience to feel like sending a text message. You press send, it arrives. That's where we're heading.
The Interest Income Windfall and Its Uncomfortable Implications
Rising global interest rates in 2022–2024 created an unexpected profit accelerator for Wise — and an uncomfortable strategic question the company is still navigating.
By early 2024, customers held over £12 billion in Wise accounts. This money — sitting in multi-currency accounts, waiting to be spent or transferred — was invested by Wise in highly liquid, low-risk instruments (primarily government bonds and bank deposits) in accordance with its regulatory requirements. As central bank rates rose from near-zero to 4–5% across major currencies, the interest income Wise earned on these balances surged. In FY2024, interest income contributed meaningfully to total revenue — the company reported it as a distinct line item, and while the precise split is somewhat blended with other income, analysts estimated that interest on customer balances and Wise's own capital contributed several hundred million pounds to the top line.
Wise chose to share a significant portion of this windfall with customers, offering interest on balances held in major currencies — up to 4.46% APY on GBP balances by mid-2024. This was smart positioning: it reinforced the brand's transparency ethos, increased customer stickiness, and encouraged larger balances. But it created a new kind of exposure. Interest income is not a product decision — it is a macro variable. When rates fall, as they inevitably will, Wise's revenue from customer balances will compress. If the company has allowed customers (and the market) to anchor expectations on interest-enhanced revenue, the reversion could be painful.
Management has been relatively transparent about this dynamic, breaking out interest income in financial reporting and cautioning that it should be viewed as cyclical. But the market, as markets do, has partially capitalized the windfall — Wise's valuation recovery through 2023–2024 was driven in part by the interest income boost. The test will come when rates normalize. The bull case is that Wise's core transfer and platform revenue will have grown enough by then to absorb the interest income decline. The bear case is that the company is temporarily over-earning, and the stock is priced for a margin profile that cannot persist.
Mission as Moat
In March 2021, the company dropped "Transfer" from its name. TransferWise became Wise — a rebrand that signaled the shift from single-product transfer service to multi-product financial platform but also, more subtly, an evolution from insurgent to institution. The provocative stunts grew less frequent. The messaging shifted from "banks are ripping you off" to "money without borders" — still ambitious, but less combative. The company's mission statement crystallized: "Money without borders — instant, convenient, transparent and eventually free."
That last word — free — is doing an enormous amount of strategic work. It implies a destination, not a current state. It frames every fee Wise charges today as a temporary imperfection, a concession to the current limitations of infrastructure and regulation that the company intends to engineer away. It is a mission statement that doubles as a pricing strategy, a talent attraction mechanism (engineers who want to work on the hardest infrastructure problems in payments are drawn to the audacity of "eventually free"), and a competitive moat — because any potential competitor that enters the market pricing at 1% or 2% is, by Wise's framing, already behind.
Whether "eventually free" is achievable or even desirable is a separate question. The most plausible long-term model is that cross-border transfers themselves approach zero cost while Wise monetizes adjacent services — account fees, card interchange, interest income, platform licensing, premium business features. This is roughly the trajectory the company is on. But the mission creates a useful tension: it pushes the organization toward constant cost reduction while forcing creative thinking about where the actual monetization will come from. Companies that live inside this kind of productive tension — Amazon's "Day 1" mentality, Costco's member-first pricing philosophy — tend to compound advantages in ways that more comfortable competitors cannot match.
The rebranding to Wise coincided with the direct listing, and the two events together marked a kind of institutional coming-of-age. The scrappy Estonian startup that hired actors to strip outside the Royal Exchange had become a publicly traded company with a £6+ billion market capitalization, regulatory licenses in 30+ jurisdictions, and infrastructure partnerships with major banks. The underwear stunts were over. The mission remained. And in the company's London headquarters — a deliberately unglamorous office in Shoreditch, where Käärmann's desk was, by multiple accounts, indistinguishable from any engineer's — the take rate continued its slow, inexorable decline: 0.57% and falling.
