Every payments conference has an AI panel this year. Sometimes more than one. Occasionally several. Most begin with the same question: how will AI transform payments?
After moderating a panel at ACI Worldwide’s Payments Unleashed Conference in London last week, I’m starting to think this maybe the wrong question for 2026.
At this stage in the technology’s evolution, it’s better to ask where is AI already creating measurable value. On this, there was remarkable agreement among the panellists. Fraud.
Whether it was Nvidia talking about Revolut’s Transaction Foundation Model, ACI discussing self-learning fraud models or PayPal describing how AI is improving internal decision making, nobody questioned that AI investment in fraud prevention is producing genuine returns today. Rule-based systems are being replaced with models that continuously learn from new transaction data, improving detection while reducing false positives.
The second area where AI is already delivering results is less obvious but equally important. Payment routing.
Using AI to decide which acquirer to use, when to retry failed authorisations or how to optimise approval rates is producing measurable commercial benefits. Unlike some of the more speculative AI use cases, fraud and routing are straightforward optimisation problems which the industry has been working on for years. Better performance is easily measured and delivers clear cash benefits.
What next? Much of the excitement today around AI centres on agentic commerce. The popular image is one of computers buying from other computers with no human involvement.
Brenden Lane from PayPal argued that’s not what we’re seeing today. Instead, AI is helping consumers decide what to buy rather than buying it entirely on their behalf. Travel and consumer electronics were the standout examples. AI can compare itineraries, explain technical specifications and narrow the choices before the customer makes the final decision. In other words, AI is improving shopping more than it is automating payments.
Erich Litch from ACI agreed. “We automatically jump to the payment,” he said. “There’s a lot of stuff that’s way more valuable in the shopping experience.”
If AI can remove friction from product discovery, supplier selection or travel planning, then the payment simply becomes the final confirmation of a decision that has already been made. The question for payment companies is whether the really changes much in their world. This could just be business as usual and, with some minor modifications, exiting card rails will work just fine.
But investment is certainly growing. Nvidia pointed to rapidly increasing AI budgets and a surge in new application releases. But there was broad agreement that agentic commerce remains at an early stage. Brendan Lane was explicit that it won’t replace traditional commerce this year, while Erich Litch compared today’s adoption levels with online banking in 2000. Interesting, certainly. Mainstream, not yet. And the lesson from 2000 is that the early innovators – remember Egg? – didn’t hit the jackpot.
Any discussion about AI can’t ignore public policy. The USA and Europe are facing in different directions with the UK rather caught in the middle. Rather than calling for sweeping new AI regulation, Lord Chris Holmes argued for technology-neutral, principles-based legislation that can survive changes in technology. It is an attractive idea, although agreeing those principles may prove more harder than writing them down. And there’s an open question whether any regulatory regime can move at the same speed as today’s product roadmaps. Sensible rules today could become dangerous tomorrow.
Less contentiously, Holmes reminded the audience that many current AI projects will fail, not because the technology is inadequate, but because they lack a clear business purpose. This is a common problem when CEOs get excited about a new technology and demand something they can put in a press release.
I don’t think the payments industry struggles to adopt new technology. The challenge is always about identifying where technology genuinely creates value and balancing investment against future returns. It’s looking increasingly like agentic commerce won’t hit prime time in the next year or two. Payments businesses might be best advised to focus investment on fraud and routing until the business case for agentic commerce is clear.
One of the more thought-provoking conversations I had recently was with Tomáš Debnár founder of Slovakian POS provider Papaya POS on a webinar I hosted for MPE.
His central argument was. “Payment is the least interesting moment in a merchant’s day.” Merchants don’t wake up thinking about card acquiring. They think about whether they’ve ordered enough milk, whether they’ll have enough staff for the lunch rush, which products are selling, and why profits were down last Tuesday.
Payments only become interesting when something goes wrong as we’ve seen with the recent Worldpay outage in the UK. Millions of merchants suddenly became very interested in payments, but only because they stopped working.
Assuming we’ve got the basics right – and that’s not a trivial ask – the real value should lie in everything around the payment.
Papaya has spent the last 13 years building software that runs restaurants and retailers in Slovakia and the Czech Republic. It started with the cash register but now manages inventory, staff, kitchens, reporting and business analytics. Every transaction feeds data back into the system.
Papaya has launched a new product called Open Market, designed to automate procurement between merchants and suppliers.
Today, many independent restaurants still order supplies through phone calls or WhatsApp messages. Orders are disconnected from the sales data sitting inside the POS system.
Papaya wants to close that loop. Its software already knows what the merchant sold yesterday. It can estimate what they’ll sell tomorrow. It can identify which ingredients are running low. The next step is for AI to recommend, or eventually place, the order automatically with the most appropriate supplier.
Tomas shows the payment becoming one step in a much larger workflow:
For me, this is a much more interesting use of AI than optimising fraud screening or generating yet more marketing copy nobody is going to read.
Tomas believes most PSPs should partner rather than attempt to build merchant software themselves. After spending 13 years refining workflows and supporting merchants on the ground, he argues that operational expertise is much harder to replicate than payments infrastructure. Papaya is already offering its AI procurement capabilities on a white-label basis for PSPs and ERP providers that want to expand beyond payments.
The story also highlights why Central and Eastern Europe is becoming an unexpectedly fertile source of merchant innovation. Countries such as Slovakia introduced mandatory fiscal reporting systems years ago, forcing even the smallest merchants onto digital cash registers. That created a rich stream of operational data long before AI arrived and catalysed a technology ecosystem that made the most of combing Android SmartPOS devices and cloud APIs.
Whether Papaya succeeds in scaling Open Market remains to be seen but it’s a potentially very interesting model for others to follow. The company is beginning with a small beta involving merchants and suppliers before expanding into Poland, Austria and Hungary.
If you’re wondering the secret behind Flatpay’s phenomenal growth, I’d recommend 30mins listening to this podcast with Sander Janca-Jensen, the company’s founder.
Key points:
The founding team are not payments people. They came from the alarm industry and realised that the same field sales model could be adapted to distribute merchant services to small businesses.
Growth is remarkable. Flatpay was founded just four years ago in Denmark but is now live in seven markets, has 100.000 merchants, 2000 staff and has hit €20bn run-rate of annual processing volume. Sander says Flatpay is taking an incremental 10% pts of the Danish market each year and 5% of the (much larger) Italian market. The fast entry into Italy is bad news for Nexi (the incumbent) and Worldline (up until now, main challenger).
Flatpay targets 1m merchants by the end of 2029 which would get it close to SumUp’s scale. Flatpay will continue opening new markets where GDP per capita is sufficient for its acquisition model to work. This means the USA will likely come before Poland.
The product set is very simple. Flatpay avoids a “long tail” of product features used by few customers. One terminal, one distribution model (field sales) and no partners to pay commission to. Field sales reps need to generate 20 deals/month and are exited swiftly if they don’t get near to this very quickly.
Sander doesn’t mention Flatpay’s innovative, although controversial, policy of offering free processing on business and international cards funded through a consumer surcharge. Competitors think this is the key factor behind its growth and one that could be replicated by others in the market.
Flatpay never places a job advert for management. All senior hires are from their network or recommended by existing staff. Sander likes BCG/McKinsey people which means we’re talking smart, hard-working people with good people skills. Sander gives them staff roles sitting next to a co-founder for 6-12 months and then moves them to a commercial director job in a new market. Every country MD has been promoted internally.
New markets are opened with a team of 10 people expatriated from Denmark. These are gradually replaced with local hires.
Global Payments’ Q1 results highlighted the remarkable impact of the Worldpay acquisition on the scale of its European business. The addition of Worldpay takes GP’s EMEA business to around €600m of quarterly net revenue – four times the level of three years ago. GP was generating just €150m per quarter in 2023. This roughly doubled following the EVO acquisition and has now doubled again. Global Payments’ European business is now larger than Nexi’s and only slightly smaller than Worldline’s.
The enlarged GP is now the largest acquirer in the UK (Worldpay’s heartland) and Poland (through eService, the JV with PKO BP inherited from EVO). It also has strong positions in Spain, Ireland and the Czech Republic, plus a toehold in Germany through its new JV with Commerzbank.
Integration between GP and Worldpay is well underway. The top four layers of the new organisation have now been defined, and the structure is explicitly global, with multinational teams organised around customer segments — SME (or SMB in American English), enterprise and platforms. Management says it remains on track to deliver $200m of additional sales and $600m of expense savings.
Genius, the POS software proposition, is central to GP’s SME strategy. Management believes that most small businesses will ultimately buy bundled payments and software, whether from GP or another provider. In the UK, the former Worldpay sales team, which previously lacked a POS software offering, is now selling Genius. Management sounded notably bullish:
“In the U.K. and Ireland, we are expanding the size of our successful mid-market and small corporate sales teams … early adoption has been encouraging, surpassing 500 locations in less than 60 days.”
Elsewhere:
“We continue to scale Genius across multiple markets such as Germany and Austria, with additional international launches planned later this year and next.”
In enterprise, there was a notable processing win at Morrisons, the large UK grocery chain, which has ended a long-standing relationship with Barclaycard. Transactions are already flowing and “we expect to have their full migration live this quarter.” GP has also won Aldi Süd’s processing business in both North America and Europe.
Other European wins announced included:
Decathlon — Spain
Lidl — Spain
ELZAB ECOPOWER, an electric vehicle charging station provider — Poland
DG Park, a car park app provider — Poland
Tescoma, a homeware retailer — Czechia
Sklavenitis, grocery — Greece
Finally, GP was also bullish about its indirect business in Europe:
“We recently expanded integrated and platforms into the U.K., where early results are exceeding expectations. Partner signings are nearly double our planned performance and feedback continues to validate a strong product-market fit.”
There are lots of big ideas in Business of Payments this month: sovereignty, AI agents, open banking and more. Meanwhile, the industry quietly gets on with consolidation, integration and the usual power struggles.
Read on for all the latest news.
The Payments Business
This month’s big news is a major strategic pivot from Adyen. After twenty years of focus on building a single-platform processor from scratch, the company has surprised the market with its first acquisition. And it’s a big one. The Amsterdam-listed payments giant is paying €750m in cash for Talon.One, a Berlin-based start-up that helps enterprises run personalised loyalty and incentive campaigns.
Adyen has long avoided M&A, citing the benefits of a unified technology platform and well known difficulty of integrating payments businesses. This deal marks a clear change in approach.
Talon claims 300 customers including Adidas, Ikea and Kingfisher but Adyen’s move is primarily about capability rather than scale. Talon’s 2024 accounts show a still-small business, with sales up 13% to €11m although it hopes to hit €60m annual run rate by the end of this year. Talon’s investors, including those who joined a $135m fundraise in 2025, will be well pleased.
Adyen believes the addition of Talon’s loyalty engine will help widen its proposition from simply optimising payments to optimising pricing: enabling merchants to tailor offers and prices to individual customers based on demographics, behaviour and spend. It may also strengthen Adyen’s position in grocery, a vertical still dominated by domestic incumbent acquirers.
Cynics might argue that Adyen is generating more cash than it can deploy internally, and that acquisitions are more appealing than returning capital via dividends or buybacks. For now, investors seem relaxed: the share price is up 12% since the announcement.
Two notes of caution. First, loyalty is notoriously difficult to execute. It’s a high-touch product and hard to integrate cleanly with payments. Mastercard tried a similar approach, acquiring SessionM for $215m in 2019 before selling it at a loss for $20m. Second, dynamic pricing risks alienating customers. Retail shoppers tend to dislike paying different prices for the same product, and the reputational jeopardy for merchants could be real.
Nexi Group is closer than ever to being taken private again. CVC is reportedly nearing agreement with shareholders of the Milan-listed payments processor, which was originally assembled through a series of acquisitions across Italy, Germany and the Nordics. CVC’s plan needs first to spin out Nexi’s digital banking unit, which operates Italy’s interbank payments network and is considered too politically sensitive to remain under private equity ownership.
Nexi was floated at €9 per share. Despite stable, cash-generative recent results which have boosted the stock in recent weeks, the share price still trades at less than half that level.
ANZ Bank is terminating its joint venture with Worldlinein Australiaand bringing merchant services back in-house after failing to find a private equity buyer. Although the move was widely trailed, the value destruction is larger than expected: Worldline paid €300m for a 51% stake in 2022; ANZ is buying it back for just €54m.
Staying in the Pacific, Worldline has also sold its New Zealand business for around €16m, a fraction of the €115m Ingenico paid for the country’s original EFTPOS network in 2018. Worldline inherited the business through its 2020 acquisition of Ingenico. One day, someone will write a book on what a mess Ingenico made of its expansion from hardware into services.
This marks the end of Worldline’s retreat from non-European markets, having now exited Australia, New Zealand, India and the US. The result is a more focused and better-capitalised business. And a more predictable one. After years of earnings misses, 2026 looks set to deliver in line with management expectations.
Mangopay specialises in marketplace payments, processing around €10bn annually. It benefits from a deep partnership with Mirakl, the French software provider behind many leading global marketplaces. Mangopay is a rare payments asset: independent and well positioned in a fast-growing, high-margin niche. Potential buyers? Adyen and Stripe already have strong marketplace capabilities but Global Payments and JPMorgan could certainly be interested.
In other corporate news, Kustom, the online checkout business spun out of Klarna last year, is spending €44m to buy Vipps MobilePay’s checkout product.Vipps Checkout serves 3,000 Norwegian merchants and processes €0.6bn in volume. Vipps runs a highly successful domestic wallet and has sensibly concluded it lacks the scale to compete with Stripe and other global players in the highly competitive market for checkout.
Euronet has bought PaynoPain, a fast-growing Spanish PSP with strong omnichannel capabilities, 3,000 merchants and €1bn in volume. Euronet, listed on NASDAQ but often overlooked by investors, already has a strong merchant services presence in Germany, Poland and Greece. The deal also brings Euronet an annual payment party.
Tab, a Swedish company that uses card-linking to send digital receipts to shoppers’ banking apps, has been acquired by Kivra, a consumer-focused bill payment service. This reinforces the view that digital receipts are a product feature, not a standalone business.
Turning to investment, it’s been a quiet month, but a few smaller raises point to where the market is heading:
London-based Ralio has raised €2.5m to help developers enable AI agents to buy goods and services.
Prague-based SoftPOS start-up Tapaya has raised €1m in a pre-seed round. Rising memory chip costs are pushing up terminal prices, giving SoftPOS a boost.
Wero, the new wallet from the European Payments Initiative (EPI), is one of two flagship efforts to reduce Europe’s reliance on American payment networks. The initiative has widespread political support. As the Bundesbank President put it: “It is not wise to outsource our sovereignty.”
So how is wero doing? The unofficial werotracker now lists almost 50 online merchants accepting wero, including Decathlon and Eventim. German merchants are starting to come on board, with several Otto Group brands live and Nexi acting as acquirer. École du Ski Français has processed the first wero e-commerce transaction, in France with BPCE as acquirer and Payplug as PSP.
This is good progress although eCommerce is the channel with the most domestic options already available. True sovereignty requires wero to work at POS but this won’t be available until next year.
Embarrassingly, wero might not be as sovereign as advertised. We learned that EPI relies on Amazon Web Services. The EPI defends itself by saying that all data is encrypted but for me, the real issue is not the risk of Trump peeking at wero transactions, it’s the risk of AWS being ordered to turn off the EPI if one or more European countries annoys the White House. Nonetheless, EPI has now woken up to the problem and says it’s “working to reduce its dependency on non-European suppliers.”
Digital Euro
The second plank of European sovereignty is the digital euro, which continues to move steadily towards a pilot launch in 2027. A key milestone this month was the decision to adopt existing European standards from ECPC, Nexo and the Berlin Group. EMVCo’s absence is no surprise, given the desire to keep US influence at arm’s length.
The agreement is good news for merchants. It suggests that accepting the digital euro will not require new hardware as existing POS terminals and webshops should only need a software upgrade. Merchants will also be happy that the Socialist Group in the European Parliament has backed calls for fees to be kept well below those of other payment methods, proposing a cap of 4 cents per digital euro transaction.
Scheming
Carte Bancaire (CB) has seen its market share fall from 90% to 75% in just five years, as newer entrants such as Revolut expand in France without supporting the domestic scheme. CB is now on the offensive, claiming that 30 issuers are preparing to co-badge their Visa or Mastercard with CB. The Gaullist sovereignty argument is always a strong one in France.
In Spain, Bizum, the fast-growing bank-backed wallet, is preparing to challenge international card dominance at the point of sale. It’s launching a contactless payment option this month, allowing shoppers to pay either via the standalone Bizum Pay app or through their bank’s mobile app. Redsys is providing the technology. Merchants are promised lower fees than with Visa or Mastercard, but there will need to be a strong consumer proposition to persuade shoppers to abandon their habits of using Apple/Google Pay.
Blik, Poland’s wildly successful mobile payment scheme, processed €104bn in 2025, up 29%. eCommerce accounted for €52bn (also up 29%), while POS volume, where Blik often operates via a virtual Mastercard, rose 30% to €10bn. Interestingly, Blik recorded 600,000 contactless transactions in Germany during 2025. Poles are beginning to use Blik outside their home country.
ISV
As software and payments converge, software vendors (ISVs) have realised they can capture a decent share of payments economics. That’s why they are increasingly pushing customers towards in-house processing. Merchants can still use third-party providers but often at a cost.
EPOS Now customers face three choices. Either continue using Dojo terminals in standalone mode but pay a penalty of up to £89 per month to EPOS Now; switch to EPOS Now Payments (powered by Adyen); or move to a new ePOS provider. Dojo is actively soliciting merchants to move to competitive products. Whether it’s EPOS Now or Dojo that wins this tug-of-war will tell us something about the new balance between software and payments
A US survey of ISVs commissioned by Adyen found that two-thirds are looking to consolidate payment suppliers. Triggers typically include expansion into new geographies or verticals, or a desire to become the primary financial system for customers, starting with payments. ISVs are likely to favour PSPs with global reach and broader embedded finance capabilities.
One consequence of the shift from ISOs to ISVs is that fewer merchant portfolios are coming up for sale. TSG, a US consultancy, notes that software vendors tend to hold onto their portfolios, unlike ISOs, where an exit has traditionally always been part of the business plan. Where assets do come to market, buyers are paying a premium for businesses with full control over pricing, churn and onboarding.
Investors sense an opportunity to roll-up small vertical software businesses, with the business case driven by monetising payments onto an in-house stack. Look at Access Group, ClearCourse, Everfield, team.blue, Vesta Software and Valsoft. This market is moving rapidly.
New shopping
Rising labour and input costs are pushing retail and hospitality businesses towards order kiosks and self-checkout machines. Recent research suggests quick-service restaurants can cut a full minute from order-to-fulfilment times. That’s impressive but customers are less convinced. Service may be faster, but it feels slower, and satisfaction has dropped across the industry.
Smart carts may offer a more pragmatic middle ground. Cameras, scanners and weighing scales sit in the trolley rather than on the shelves. In Belgium, Colruyt seems happy with its pilot. In one store, 10% of shoppers choose a smart cart and many like them so much they will queue at peak times for one to become available.
Smartcarts are expensive – I’ve been told €1,000-€2,000 each depending on individual retailer requirements – but are expected to pay back the investment with advertising income. This could be as much as $210/month each, according to analysts following Cust2Mate, an Israeli business listed on NASDAQ Agentic Commerce
Agentic commerce
Agentic commerce – where an AI buys goods or services without a “human in the loop” – is attracting a lot of attention. But consumers are not yet ready to hand over control. Research shows they are comfortable with AI making recommendations, but reluctant to let it complete transactions. The perceived risk of mistakes remains too high.
Merchants, by contrast, are moving ahead. Ravelin finds that 75% of large merchants are already integrating one or more agentic commerce protocols. Interestingly, they trust AI customers (marginally) more than human shoppers. Their concerns are less about mistakes and more about fraud – specifically, how to authenticate and monitor non-human actors.
Some merchants may not be real people. An AI opened a store, bought the stock, hired the staff and did the PR. In San Francisco obviously.
The industry is beginning to respond. American Express has launched an Agent Purchase Protection programme, under which cardholders are indemnified against charges resulting from AI agent errors, provided both the agent and the merchant are approved by Amex. This is an early attempt to build trust on the consumer side.
Further down the stack, acquirers appear largely unconcerned to the point of complacency.Research from Visa suggests that “nearly all acquirers are confident that agentic commerce can be supported through enhancements to existing infrastructure.”
Fraud
Worldpay has lost its appeal at the French Supreme Court, following a 2022 judgment. The company had been held liable for failing in its duty of vigilance when it processed transactions for an intermediary linked to a series of fraudulent forex trading sites.
This is a civil case, not a criminal one, and the damages are relatively modest (around €50,000), reduced by the court due to the victim’s own contributory negligence. That said, the principle is important, and the judgment may have wider implications. Payment processors are not neutral pipes. If you enable payment flows linked to clearly problematic activity, you may be held responsible – not just for your direct customers, but for their customers too.
It’s not just e-commerce fraud that PSPs should be worried about. An excellent BBC investigation shows the extent of criminality on Britain’s high streets. Convenience stores are openly selling drugs over the counter and accepting card payments. Payment providers need to be all over this, or regulators will act.
Meanwhile, Visa’s tougher fraud monitoring regime, the VAMP programme, has finally gone live but many merchants are still not ready. Coinflow has a helpful explainer. The outcome is likely to be stricter enforcement of strong customer authentication in Europe while merchants will have a stronger incentive to steer riskier traffic away from cards and onto A2A rails.
Open banking
There were 31.3m open banking payments in the UK during March. The rate of increase continues to decline and the number of transactions was just 37% higher than a year ago. This is far from the explosive growth predicted for A2A payments.
Many in the industry are pinning their hopes on variable recurring payments (VRPs), the open banking equivalent of direct debits. These should gain broader adoption later this year as standards are established, initially for charity donations and utility bill payments.
Some research from GoCardless shows that merchants believe VRPs can certainly fix some of the problems the experience managing recurring card payments. But the issue for PSPs and their investors is that there’s very little money to be made in processing direct debits, even feature-heavy ones like VRPs. This is clear from GoCardless and its 2025 results.
GoCardless bought Nuapay, a direct debit and open banking vendor, in 2024 for £28m. In 2025, Nuapay generated just £7.7m of revenue from £27bn of volume. Direct debit processing is a technical service: the provider doesn’t take risk and normally can’t charge an ad valorem fee. The result is stable, low-margin revenue, but little excitement as seen in the impact of Nuapay on GoCardless’s overall take rate.
Overall, GoCardless made an operating loss of £25m on revenues up 18% to £155m. Mollie, the well-capitalised Dutch PSP, has agreed to buy GoCardless for €1.1bn – a very generous deal for GoCardless shareholders.
Business lobby groups oppose the move but as the public uses cash less often, the cost of handling it is rising. Even so, estimates still suggest cash remains marginally cheaper than debit cards. Germany’s Bundesbank puts the cost of accepting cash at 43 cents per retail transaction. In Britain, the BRC has a similar estimate of 0.15% of value, that’s around 45p on a £30 purchase.
In Holland, a new initiative will allow shoppers to withdraw €100 notes from ATMs in supermarkets. This could be helpful, but since using such high-value notes is unusual, one commentator observed: “You are considered a criminal if you want to pay with it, right?”
Crypto
While Fintech commentators fizz with excitement about stablecoins, there’s little evidence of these new currencies being used for retail payments. New research from the Kansas City Federal Reserve (below) shows stablecoins remain primarily a tool for moving into and out of cryptocurrencies. Less than 1% of stablecoins are used for payments. This is backed up by a new paper from Visa’s Economic Empowerment Institute, which concludes that “stablecoin activity represents a small share of everyday economic transactions”. As Andrew Dresner puts it, “everyone should just calm down.”
Figure 1 Source: Kansas City Federal Reserve
Some in the US administration hope that explosive growth in dollar stablecoins – essentially tokenised US Treasuries — will allow it to continue running large fiscal deficits. European governments are worried and have proposed that the EBA should have a “kill switch” that bans dollar stablecoins if an issuer is found to be acting against EU token holders’ interests. This could get messy.
Mastercard has cut emissions by 44% since 2016, mainly through moving data centres to renewable energy. It’s good to see Mastercard still caring about the planet when many others seem to have given up.
And finally
Apple’s Eddy Cue explains how he sold millions of songs at 99c without losing all the margin to credit card processing fees.
Where to find me
I’ll be at Money 20/20 Europe in Amsterdam on 2/3 June and at ACI’s Payments Unleashed in London on 29/30 June.
I work with a small number of companies on speaking, moderation and advisory roles when my diary allows. Feel free to get in touch.
Forrester’s latest research report underlines the growing capability gap between the three leading global processors – Stripe, Adyen and Checkout – and the rest of the industry. Nuvei is still a contender, Worldpay gets praise for the quality of its relationship management and Paypal remains a viable option for the US. Nobody else comes close.
The analysts find that enterprise buyers are showing an increased focus on the quality of relationship with their processors and the pace at which vendors can add new capabilities. High acceptance rates, low prices and rock-solid delivery are a given. Forrester reports are normally quite expensive but Adyen has done a deal to make this one available free of charge.So has Stripe. So has Checkout. You can take your pick.
Meanwhile, in legacy land, Worldline is slowly recovering from last summer’s near-death experience. It has raised €392m from three French banks as part of a proposed €500m capital raise. This includes further investment from Crédit Agricole, giving fresh impetus to CAWL, the bank’s joint venture (JV) with Worldline which has got off to a quiet start.
12,000 miles away, ANZ Bank is looking to unwind its JV with Worldline. Worldline paid €300m for 51% of ANZ merchant services in 2022, gaining 80.000 merchants and 20% of the Australian market. ANZ Worldline lost c.€40m in 2024 and required additional capital from its parents. Managing a troubled JV with an unhappy partner is a major distraction for Worldline’s management in Paris. A deal looks likely soon.
Nexi, Europe’s largest processor by volume, has lost its CEO after a capital markets day sent the stock down 16%. The business model is, quite sensibly, shifting from M&A-led growth to cash generation but investors remain unconvinced. Nexi is offering higher dividends, with the stock now yielding a generous 9%.
Neobanks are moving into payments. Revolut’s merchant acquiring offer, based on a Newland terminal, is now live in 19 countries, ten more than last year. Volumes tripled in 2025, though no figures were disclosed. Wise has also entered acceptance, offering payment links. Europe/UK fees are 1% for domestic consumer cards and 2.9% for others.
PayPoint, the UK’s leading ISO, said its Handepay subsidary will stop selling to small business customers, citing high churn and low margins. This reflects competition from the “tap pack” – SumUp, Viva.com, myPOS, Square, Dojo, Flatpay etc – which offer modern technology and digital services. Legacy ISOs like PayPoint and takepayments (now part of Global Payments) have had to cut prices to compete.
Flatpay will soon face competition at home market of Denmark. Unzer is launching POS acceptance via its Quickpay and Clearhaus businesses. These already serve 30,000 online merchants. Verifone will supply hardware. Unzer has also acquired AllCash’s POS operations in eastern Germany, adding 500 terminals and 2.5m transactions.
Nayax, the Israel-based unattended payment specialist, reported strong 2025 results, with revenues up 28% to $400m. European sales rose 21% to $92m and UK sales also grew 21% to $47m.
In corporate news, Vipps MobilePay is selling Vipps Checkout to Kustom, a Klarna spin-out, for NOK 490m (€44m) including 3,000 merchants in Norway and NOK 7bn (€0.6bn) in volume. This looks smart for both sides. Vipps exits a business that cannot compete with global players, while Kustom gains access to the Vipps ecosystem as a preferred partner. Despite the rather high €14,000 per merchant price, distribution is likely to be the real prize.
Outpost, a London-based merchant of record vendor, has raised €15m to support cross-border commerce. Outpost takes control of tax, logistics and sales for merchants in each market.
Tipjar which offers a cashless tipping solution, has raised a further £4.5m and £11m in total. It now serves 5,000 UK sites, processing £130m of tips annually and has launched in the USA and France.
Finally, an intriguing US court case: Block is suing London-based Enigmatic Smile, a card-linked loyalty provider, for $557k in unpaid commissions. Enigmatic Smile doesn’t dispute the sum but still isn’t paying up. Block is seeking punitive damages.
Conference round-up
This year’s MPE conference in Berlin showed Europe’s payment industry in good form – record delegate numbers, buoyant sponsorships and plenty to discuss. Here’s my three-minute report.
I moderated the panel on composable (modular) payments joined by Mastercard, Payrails and Kingfisher. The conclusion: merchants with complex needs will look to build or rent their own payment stack. Those with simpler requirements – even though volume may be very large – will stick with full-stack PSPs. Read more on the Business of Payments blog.
Back in London, my take after chairing a Pay360 discussion on payment data is that payment data is great for fixing payments, but not necessarily useful for much else. Read more on the Business of Payments blog.
This was reinforced by President Macron in a video message to a Carte Bancaire conference in Paris. He backed wero as an alternative to Visa and Mastercard: “It’s about controlling the security of our trade, the continuity of our economy, our ability to decide for ourselves, and therefore our independence.”
The French financial establishment is particularly worried that 30m cards are branded solely Visa or Mastercard, without Carte Bancaire co-badging for domestic use. Carte Bancaire, which has been losing share, has begun to modernise and believes it has turned a corner with 30 issuers and acquirers joining the scheme. Macron will welcome BPCE’s decision to co-brand all cards by 2027, with the bank explicitly citing payment sovereignty.
Mastercard has felt compelled to respond. Kelly DeVine, European President, reaffirmed the network’s commitment to investing in Europe and complying with local laws, adding that Mastercard would challenge in court any unwarranted attempts to disrupt its ecosystem. The problem for Mastercard is that nobody believes this. If instructed by the US government to block an individual, merchant or country, the card schemes would have little choice but to comply. Hence the growing political momentum behind credible European alternatives.
The unofficial werotracker shows 51 banks having gone live with for P2P or C2B transactions. 30 webshops can accept Wero for eCommerce transactions with a further 35 merchants announced but not yet live.
One reason for optimism is Wero’s flying start through incorporating iDEAL, the highly successful Dutch online bank transfer system. Wero operates more like a traditional scheme. This brings helpful new product features such as pre-authorisations and subscriptions but wero also includes chargebacks which iDEAL does not offer. Dutch merchants are very unhappy about having to take responsibility for credit risk.
In Euroland, momentum is growing behind the digital euro. The ECB is inviting PSPs to take part in a pilot during the second half of next year. This would involve central bank staff and a few selected merchants such as the ECB staff canteen. Full launch is scheduled for 2029. Although sovereignty has become the key rationale for the digital euro, merchants will be hoping fees will be cheaper than cards. The German retail federation is demanding transactions are priced at 4c maximum. French retailers have proposed 0.1% up to a maximum of 4c.
Mastercard has announced plans to spend up to $1.8bn buying BVNK, a stablecoin infrastructure provider that holds, moves and converts stablecoins into real money and back again. None of that capability is clear from the corporate video.
Mastercard’s acquisition record is mixed. The network is reportedly selling the Nordic bill payment businesses it bought from Nets. Today’s value is likely much lower than the $2.9bn Mastercard paid in 2021 even though the two businesses – Betalingsservice in Denmark and AvtaleGiro in Denmark – are highly profitable, generating $100m EBITDA from $370m revenues.
Giro is fighting back with new business model in which network operators will pay a “scheme fee” of c.2 bps for the first time. In return, the acceptance side of the payment industry is promised new product features and a formal role in the governance of Giro. The scheme is also emphasising payment sovereignty: “Bezahalen. Made in Germany.”
Irish banks have finally launched Zippay, an instant payment service allowing users to send money to other domestic bank accounts from their mobile banking app. While marketed as P2P, experience elsewhere suggests that small businesses quickly start using this kind of product to accept money from the public.
Nexi is supplying the technology although it’s not clear why the Irish banks have built their own service rather than joining Wero.
What’s on your terminal?
Designed well, a payment terminal’s screen will speed transactions, keep merchants happy and reinforce the supplier’s brand values. Done badly, both shopper and retailer can be easily confused.
Research presented by Flagship Advisory at MPE suggests software vendors (ISVs) are now the main distribution channel for SME payments in both the USA and UK. The bank channel, almost extinct in the UK, still accounts for 48% of POS distribution in Europe but ISVs are becoming an increasingly important sales channel for processors everywhere.
Agentic commerce
It seems 2026 will not, after all, be the year in which AI agents start making meaningful inroads into eCommerce. Apologies for the confusion but this technology is moving through the hype cycle at record speed and we’re already entering the trough of disillusionment. Andrew Dresdner has a good explanation of why we’re hitting some speed bumps.
This matches feedback from merchants at MPE. Everyone agrees agentic commerce will be big but it’s clear that the technology is not quite ready. The industry needs to be making adoption much easier. The alphabet soup of protocols (sometimes outnumbering transactions) is not helping.
Adyen has a issued a good report which outlines the infrastructure challenges ahead including merchant product data and checkouts designed for humans and fraud models set up to stop bots, not encourage them.
Venture capitalists at A16z believe this “headless” commerce as the future; one in which merchants exist as product, pricing and promotion files without the need for a website or brand marketing. Interestingly, A16z no longer sees stablecoins as the likely payment solution for agentic commerce. Mastercard and Visa have got their act together and “the rails are no longer the bottleneck.”
Agents will most likely pay with a virtual card and there are a growing number of start-ups looking to supply this need. AgentCard will give your agent a debit card with spending limits. There are many others.
Symphopay, based in Bucharest, has a rare capability to dynamically route card payments at point of sale. Romanian acquirers offer generous on-us pricing which means that merchants have an incentive to take a terminal from each domestic acquiring bank. Symphopay collapses this chaos into a single device and supplies 3000 terminals to 22 local retail chains including a recently completed 1750 store roll-out for Profi. Symphopay has built a full service with modest capital raise of just €1.4m and reportedly made a small profit in 2024 on turnover of €1m.
Open banking
British consumers made 33m open banking payments in February. While that’s 39% up on the previous year, it’s well below the growth rate needed to sustain the industry’s business model. The public don’t yet seem interested. A new survey from Yaspa, an open banking vendor, shows only 38% of UK adults familiar with the term “Pay by Bank.”
The industry’s hopes are pinned on commercial variable recurring payments (cVRPs) – open banking’s initiative to improve direct debits. Yapily has a good explainer. The first wave of cVRPs – for utilities and charities – should go live in May. To help consumers understand the new product, vendors such as Truelayer are beginning to rename cVRPs as “bank on file.”
Farewell Vibe Pay, a high-profile London-based open banking start-up. Despite its celebrity backers, including the Reform UK party treasurer, investors concluded that free payments is a challenging business model. The business will be liquidated and the ten remaining staff lose their jobs. A deal to sell Vibe Pay to Banked, a very well-funded open banking player, was announced last year but failed at the due diligence stage.
SoftPOS
2026 looks like the year when SoftPOS goes mainstream.
Munich-listed Rubean AG reports revenues doubling in 2025 to €3.8m, mainly from one-off licence sales to a growing rosta of enterprise merchants in Spain, UK and Eastern Europe. Delivery services have been early customers but Rubean also announced a good win with the RAC’s roadside assistance patrols in UK and Ireland.Fiserv is the acquirer. The momentum continues: Rubean’s sales are expected to grow “high double digits” this year.
Worldline says Tap on Mobile (its SoftPOS product) is now deployed in 23 markets, processing €760m volume in 2025 – still relatively small but three times 2024’s number.
Industry sources suggest that Dojo is looking to standardise its SME terminal proposition on SoftPOS using using MyPinpad’s software application. If true, this would be a major boost for Licentia Group, based in Cardiff and the parent of MyPinPad. Licentia lost £2.7m in the year to June 2025 on turnover of £4.6m.
Policymakers are grappling with the future of cash. As digital money becomes more prevalent, there are growing concerns around financial inclusion and economic resilience.
But this is only heading one way. As cash usage declines, the cost of cash handling becomes expensive and fixed costs are spread across fewer transactions. This pushes up the cost of operating ATM networks or accepting cash in shops. In Poland, Euronet has started limiting withdrawals using Blik in protest against the low fees it receives from the customer’s bank.
FXC has published a well-researched buyers guide to cross-border stablecoin products. I’ve come to two conclusions. Firstly, the sector is still very immature and the value chain is bafflingly complicated (see below). Secondly, the consumer to business (ie, merchant payment) opportunity is limited. Bottom line: Europeans are unlikely to be using stablecoins to buy from European retailers in the foreseeable future.
Figure 2 Stablecoin value chain – source: FXC
For the moment, stable coins will be primarily used to support global treasury flows to and from countries with either high inflation or limited access to dollar bank accounts such as Nigeria or Argentina. BVNK’s recent consumer research backs this up.
One problem with these markets is that KYC is difficult, meaning that most of the cost lies in compliance rather than the rails themselves. Stablecoins don’t make KYC any easier and screw-ups can be expensive. Bridge, owned by Stripe, has some explaining to do after shipping 12 Mitsubishi trucks to sanctioned Chavez family members in Venezuela.
One of the reasons for SoftPOS growth is that standard payment terminals are becoming more expensive. Memory chips are up 500% since 2025 according to the CEO of Ingenico.
The divergence between new and legacy acquirers is explained by Dwane Gefferie. If you are not offering routing, smart retries, signal enrichment and network tokenisation, you’re not in the game.
For the second successive year, I moderated a panel on modular payments at the MPE conference in Berlin. We are tracking the shift in enterprise buying behaviour for payments from taking a full stack service from a single PSP to something that increasingly looks like a mix and match set of cloud infrastructure.
Maria Parpou (EVP, Merchant Cloud, Mastercard) explained that managing payments is becoming more complex – more PSPs, more schemes, more territories, more channels and more alternative payment methods – while profit margins are declining in many retail categories. Her view was that any decision about modularity should be driven by outcomes, not architecture. For merchants, the priorities remain constant: availability, authorisation rates, and helping the business grow. Mastercard’s Merchant Cloud is one response: a composable platform offering acquiring, gateway, tokenization and fraud services under a single contract, with the flexibility to add 3rd party partners.
Chris von den Hoff (Enterprise Sales, Payrails) took this further, arguing that modular infrastructure is a prerequisite for “intelligent payments.” The real blocker today is fragmented data across multiple PSPs and legacy internal systems. Payrails positions itself as the connective layer, normalising data, enabling orchestration, and increasingly automating operational workflows like reconciliation and chargebacks. Chris explained that merchants don’t ask for modularity – they ask Payrails to help fix approval rates, reduce costs, or expand internationally.
From the merchant side, Fiennes Davy (Group Head of Payments, Kingfisher) offered a useful counterpoint. Kingfisher, despite its scale, is prioritising simplification over orchestration. With regulatory changes like PSD3 looming, the choice is often binary: take on the complexity of getting as license – especially for any merchant running a marketplace – or outsource to a smaller number of strategic partners. It’s clear that for many merchants, modularity is not an immediate priority.
The conclusion? The stack will remain mixed. Some merchants will consolidate with a single supplier that does everything. Others will opt for a set of modular services, some built in-house, others delivered from the cloud. I suspect the real battleground will be who controls the data.
At Pay360, I moderated a panel on payment data – we’re talking about the information contained within and the meta data surrounding a transaction. While, payment data is becoming central to how businesses operate, there’s a risk of overstating its role.
Steve Bisoffi of Revolut was pretty clear: “No viable, sustainable business can operate without detailed payments data.” At Revolut, payments are monitored continuously across dashboards tracking growth, fraud, disputes and performance. Without that visibility, “you don’t really understand how your business is operating.”
I agree but only up to a point.
Payment data is extremely powerful when it comes to improving payments themselves: acceptance, fraud, routing, cost and customer experience. But it is not, and cannot provide a complete view of the business.
Merchants’ real systems of record sit elsewhere – in CRM, ERP and ePOS systems. That’s where you find SKU-level detail, customer identity, promotions and margin. Payment data, by contrast, tells you where and when a transaction happened, but not what was bought or why.
Much of the discussion on the panel focused on using payment data to drive wider business outcomes. Sophie Wadsworth of BNY (Bank of New York) described how payments have shifted from cost centre to revenue driver, with use cases ranging from cashflow optimisation to reducing churn from failed payments.
But even here, the value of payment data is often indirect. It improves conversion by reducing failure rates. It improves cash visibility. It supports forecasting. It is a helpful complement to core commercial data but of limited value on its own.
Where payment data really is valuable is in fixing payments. And fixing payments can deliver quick improvements to any merchant’s bottom line.
Phil O’Hagan of Sainsbury’s gave a good example. Analysis of re-authorisation behaviour showed customers repeatedly triggering new payment requests when adding small items to their basket. By adjusting tolerance rules, Sainsbury’s reduced friction, cut contact centre calls by around 2.5 million, and saved a lot of money.
This is best practice payments optimisation: small changes, driven by data, delivering outsized operational impact.
The same applies at scale. Mastercard’s John Wheeler said the network was using data to improve acceptance rates and guide acquirers. Tokenisation, he suggested, could reach near 100% globally by 2030, another example of data driving better payment outcomes. Yes, payment data can help merchants decide where to open stores or better understand share of wallet, but these use cases remain relatively niche.
For merchants, the actual challenge is integrating payment data with their systems of record and those of their partners.
David McGuiness of FoodHub highlighted the importance of linking payments with POS data so that “they speak to each other.” Only by combining these datasets do you get a complete picture of the business. He also highlighted the challenge of getting FoodHub’s restaurant partners to take advantage of the intelligence offered.
This integration challenge has important implications for vendors.
First, data is now a core part of the product. PSPs cannot compete on price and acceptance alone. The quality, granularity and accessibility of data are becoming key selection criteria and, merchants tell me, increasingly reasons for churn. A very large merchant (not represented on this panel) told me his acquirer still provides management information via a monthly spreadsheet and invoicing on a pdf.
Second, generic dashboards are no longer enough. As Charles Aji of Collinson noted, fragmented PSP data creates internal friction and “revenue leakage” when issues are spotted too late. Many merchants are responding by building their own data layers, aggregating data across PSPs and enriching it with their own systems. With the adoption of AI coding tools, this is now much easier than it used to be. Charles had expected to buy a payment dashboard from an established vendor but his tech team has done what’s needed at a fraction of the cost.
If merchants like Collinson are building their own data capability, PSPs will find it ever harder to demonstrate the competitive differentiation needed to escape being just a pipe connecting merchants to schemes.
Finally, there is a growing expectation of real-time, actionable data across the ecosystem. Steve Bisoffi called for greater transparency from Visa and Mastercard, particularly around fees and data quality issues. Better data sharing could improve acceptance, reduce fraud and enhance customer experience.
Payment data is increasingly critical to making payments work and increasingly central to how merchants choose their partners. Vendors need to get the basics right, but not oversell the ability of payments data to do more than make payments better.
2025 results from the leading payment players show continued consolidation. The large, modern processors are taking share from incumbents and, more importantly, winning the fast-growing merchants of the future.
Stripe’s volume grew 34% in 2025 to $1.9 trillion – an astonishing $500bn of incremental volume. Even so, that is equivalent to just 2% of global Visa and Mastercard volume, so Stripe has plenty of room to grow.
The merchants onboarded last year are Stripe’s fastest-growing cohort ever, which management suggests could reflect “a larger inflection in entrepreneurship and creativity facilitated by advances in large language models.” Yes, it’s AI again already.
Stripe’s CEO took to Youtube to read his annual shareholder letter out loud. It’s twenty minutes well spent.
Simon Taylor explains why it would be a great deal for Stripe – something involving stablecoins that was beyond me but worth a read if you understand these things. A private equity bid seems more plausible: sell Venmo and Braintree and run the remaining PayPal business for cash.
Checkout.com also reported a monster year, with volume up 64% to “over $300bn.” According to the CEO’s annual letter, the company is now EBITDA positive “with no adjustments.” Checkout remains private and offshore, so the detailed numbers are not public. Unlike Stripe, Checkout focuses exclusively on enterprise merchants, including 63 customers each processing over $1bn, up from 39 in 2024.
Like many PSPs, Checkout is leaning heavily into AI. Internally, AI now automates 100% of rejected transaction distribution and is expected to handle 50% of support volume by 2026. Customers will also be reassured to learn that a “proprietary hierarchical contextual multi-armed bandit algorithm” decides how transactions are routed for the best chance of success. Every PSP needs one of those.
Back in the real economy, Adyen continues to outperform its omni-channel competitors. Reported 2025 volume was hit by the loss of Cash App processing in the US, but the underlying business remains very strong. Revenue grew 21% to €2.4bn, with EBITDA margins still above 50%.
Adyen is generating serious cash – €1bn in free cashflow with €5bn sitting on the balance sheet – and may soon have to choose between acquisitions or paying a dividend. Fintechs hate paying dividends; it makes them feel old.
Adyen remains the go-to vendor for multi-country, multi-channel merchants. It recently rolled out terminals across 943 Starbucks stores in the UK, Austria and Switzerland in just seven weeks, impressively completing the migration during working hours. The terminals support store-and-forward, allowing transactions to process (at Starbucks’ risk) even if connectivity drops. Adyen says 45,000 sales have already been saved in the first few months of operation
In Europe, Adyen’s revenue is catching up with the two major incumbents – Worldline and Nexi – and is now slightly ahead of Global Payments.
Nexi reported solid results. Merchant solutions revenue rose 2% in H2 2025, with strong growth in Germany offset by merchant losses to Worldline in Italy and ongoing e-commerce weakness in Poland. Like Adyen, Nexi has surplus cash. Unlike Adyen, it is happy to return value to shareholders, increasing its dividend by 20% – a 9% yield at today’s share price
At Worldline, the spring clean continues. After selling its North American operations and PaymentIQ, the company has now divested its Indian business to BillDesk for €60m. This looks sensible – Worldline was unlikely to achieve scale in India and the proceeds help strengthen the balance sheet.
But the 2025 numbers are still painful. Merchant services revenue fell 1.4% to €3.2bn, while adjusted EBITDA dropped 20% to €624m. There are some early positives from the turn-around. Payment volume grew 3% in Q4, with growth returning in the Nordics, Germany and Switzerland. Worldline also announced a further €290m write-down on its Ingenico preference shares, now valued at zero.
Figure 1 Source: Reuters
Global Payments reported its final quarter before consolidating Worldpay’s operations. European revenue reached €300bn in Q4, up 6%, with strong contributions from the former EVO businesses in Poland and Greece acquired in 2023.
Global has begun integrating its European operations with Worldpay, which brings a strong enterprise e-commerce customer base but has been losing share in its UK SMB heartland to Dojo and others. Much will depend on the successful rollout of Genius – Global’s Clover competitor – across UK and European SMB channels, starting with Germany and Ireland.
Dropping down a tier, VR Payments is one incumbent gaining market share. The payments arm of the German co-operative banks reported sparkling 2024 results. Volume rose 38% to €6.1bn across 312,000 terminals. Net revenue jumped 252% to €73m, while net profit increased from €2m to €7m.
Also in Sweden, Swedbank has separated its merchant services division as – guess what – Swedbank Pay, serving 65,000 merchants across Scandinavia.
Meanwhile the story around Guavapay is becoming clearer. Mastercard was reportedly pursuing £11m millions in unpaid debts, while the UK FCA had identified that around a whopping 10% of its transactions were scams. See below. Blimey.
In corporate news, Corpay is selling PayByPhone for an undisclosed amount just two years after acquiring the business from Volkswagen for about $300m. PayByPhone operates a parking payment app across the UK, France, Switzerland, Germany and North America, generating roughly $100m in annual revenue. However, its consumer focus increasingly sat outside Corpay’s B2B payments strategy.
The team at Reward will be delighted with a $230m cash exit to Dan Wagner’s Rezolve AI. Reward is the only European company to turn card-linked offers into a profitable model and has strong relationships with banks and retailers in the UK. Even so, loyalty is notoriously difficult to scale. After twenty years in business, Reward generated £2m operating profit on £53m revenue in FY24.
That price represents a striking multiple for Reward’s shareholders. Rezolve’s long-suffering investors, whose stock has fallen from $10 at flotation to $2.70 today, may have questions. If you can make sense of Rezolve’s press release, do let me know
Turning to embedded lending, Flowpay, the Czech merchant cash advance (MCA) provider, has acquired Tapline in Germany (terms undisclosed). Tapline provides short-term lending to subscription businesses secured against recurring revenues from platforms such as Stripe, Recurly, Chargebee and QuickBooks. This complements Flowpay’s more retail-focused integrations, including Shoptet and Shopify.
Expect a blockbuster IPO in Warsaw as Blik prepares for a €2bn flotation. The mobile payment standard now has more than 20m active users in Poland and accounts for 60% of online transactions by value. Blik’s net revenue equates to roughly 0.17% of transaction value (around €0.06 per transaction). The system is owned by a consortium of Polish banks, with Mastercard also a key shareholder.
In fundraising news APEXX Global, the London-based payment orchestrator, is raising a further $10m. Growth will need to accelerate to justify the additional capital, but management points to new travel clients including Jet2, Iglu.com and Norse Atlantic. 2024 results show an operating loss of £6.9m on revenue up 21% to £5.7m. Merchants will be interested to note that 38% of APEXX revenue comes from commission payments from acquirers.
Silverflow, the Amsterdam-based provider of outsourced card processing for fintechs, PSPs and smaller acquirers, has raised $40m. The company plans to increase headcount from 85 to 120, focusing on product development and expansion in the US and Southeast Asia. Management claims almost $100bn annually processed for customers including Buckaroo, Bolt and payabl. Latest accounts show 2024 revenue tripling to €3.2m, so there is clearly still a long road ahead.
The payment world is gathering in Berlin next week for the 2026 MPE conference and the organisers are offering a small discount for readers of Business of Payments (see below).
This year’s start-up competition looks particularly interesting. I’ll be rooting for Inqyre, which uses AI to orchestrate onboarding and scheme compliance for ISOs, PSPs and acquirers. Inqyre is currently building its MVP in partnership with large PSP and expects to begin onboarding customers later this year. I’ve made a small investment. It’s my first and I’ll let you know how we get on.
America’s unexpected attack on Iran has reinforced concerns about European payment sovereignty. Politicians increasingly believe the continent is too dependent on an unreliable ally.
Europe’s response is threefold: the rollout of wero, the payment wallet developed by EPI, a Franco-German-Benelux consortium of banks; the digital euro, expected from the ECB around 2029; and continued investment in local debit schemes such as Girocard and Carte Bancaire.
British banks are now uneasy. “If Visa and Mastercard were turned off, it would send us back to the 1950s,” one told a newspaper. The banks see an opportunity to regain control of domestic payments and are reportedly considering launching a competitor to Vocalink. Visa and Mastercard are expected to participate, so the initiative may prove less revolutionary than it sounds.
#werowatch
Wero, the mobile payment service built on SEPA Instant, is starting to gain momentum. The scheme launched with 50m users, largely inherited from iDEAL (Netherlands), Payconiq (Belgium) and Paylib (France), although Wero has already added 8m German users of its own.
Bank support is also strengthening. Commerzbank has rejoined Wero, having walked out in 2022 when the project still planned to build an expensive new card scheme. Many banks, including the German Sparkassen, are now actively promoting wero.
According to the unofficial werotracker, 17 merchants are now live – up from six last month – with another 29 announcing plans to accept Wero. The platform’s 50m users can use wero online at Ahold Delhaize, the first Belgian retailer to support the scheme. POS payment should follow in 2027.
Dutch merchants, accustomed to some of the lowest payment fees in Europe, are less enthusiastic about wero’s ad valorem pricing as it replaces iDEAL. Wero does, however, introduce features previously unavailable with iDEAL including subscriptions, chargebacks and pre-authorisations.
Farewell Lyf Pay. Despite backing from the French establishment – BNP Paribas, Crédit Mutuel, Groupe Casino and Auchan – the QR-based payment and loyalty wallet never gained meaningful traction. After burning through at least €180m, the project is being shut down. Meanwhile, a leading Australian retailer has also given up on QR code payments saying “shifting consumer behaviour in a heavy tap and pay market to scanning a QR code is hard and real blocker of true future scale.”
The lesson is familiar. Building a beautiful payment product is easy; getting people to use it is much harder.
ISV
Software vendors have become valuable distribution partners for payment processors. ISVs typically take 30–50% of net revenue but churn is very low, keeping customer acquisition costs down for everyone.
It is probably too early to panic. The predicted “SaaS-pocalypse” has yet to arrive. Take Shopify, the world’s largest eCommerce platform for small businesses. Merchants processed $248bn through Shop Pay in 2025, up 37% year on year. Shop Pay runs on Stripe.
Retail and restaurant software vendors that supply hardware have an additional switching barrier.
One example is Epos Now, the UK-based retail and hospitality ISV. The company reported another strong year in 2024/25: revenue up 35% to £113m and operating profit up 53% to £10m. More than half of sales now come from Europe, with strong growth also reported in the US and Asia-Pacific. The company recently paid its founder a £3m dividend and is hiring 150 additional staff in Norwich to support expansion.
There is still significant room for payments growth. Around 80% of Epos Now revenue comes from software subscriptions, compared with 33% at Lightspeed Commerce and 15% at Toast. That makes Epos Now an obvious M&A target for the large processors.
Although many bundles look similar, software companies that sell payments command much higher valuations than payment companies that sell software. Andrew Dresdner highlights Toast and Shift4. Both process similar volumes and focus on restaurants, yet Toast, the native software platform, is valued at roughly three times Shift4.
New shopping
The demise of Amazon Go probably says more about Amazon’s convenience-store merchandising than about the future of the technology. Autonomous stores are thriving where they make retail sense, typically as walk-in vending machines rather than full supermarkets.
In Poland, Żabka Nano already operates around 50 such stores. No app required. Simply tap your payment card at the entrance.
In the US, VenHub has gone a step further. Customers order through an app that geo-locates them, and robots prepare the order just as they arrive at the store. VenHub says it costs about $150k to build one of these units, which sell for $275–300k. Nice business.
Agentic Commerce
No technology is moving faster than agentic commerce, and it is becoming increasingly difficult to keep up with the announcements and new acronyms. The big unanswered question for payments is who controls checkout when agents start making purchases.
Stripe has helpfully mapped five levels of agentic commerce. To stay on trend, I asked ChatGPT to turn the framework into a graphic (see below). Stripe believes we are currently somewhere between stages one and two, although it admits that “there’s no forecasting exactly where agentic commerce will be by the end of 2026.”
Adyen’s management, in typically down-to-earth Dutch style, said: “At the moment the number of agentic transactions is still immaterial on our platform… it’s not going to drive short-term revenues… and it’s not a big part of our 2026 expectations.”
Shopify reports that shoppers are experimenting with AI-powered search, but only about a dozen of its 6m merchants are actually selling through AI tools so far. Shopify says the bottleneck lies with the AI platforms themselves, which have been slow to build usable merchant tools. OpenAI, for example, currently manually onboards each Shopify merchant.
As a result, OpenAI has reportedly abandoned plans for direct checkout inside the chat interface. Instead, transactions will be routed through third-party retail apps connected to its platform.
Agentic commerce may eventually transform retail but for now it looks more like AI-powered product discovery than AI-driven purchasing.
Openbanking
Growth in UK open banking payments slowed again in January. Around 30m payments were made, up 40% year-on-year. Even so, 2026 does not yet look like the year of open banking. Pay-by-bank still struggles with low consumer awareness, lack of retail bank support and limited merchant incentives compared with cards.
Many in the industry are pinning their hopes on VRPs (variable recurring payments) — the open banking equivalent of direct debit. These should get a boost as GoCardless, the UK’s largest direct debit provider (and soon to be acquired by Mollie), recently won a contract with Octopus Energy covering 5.5m mandates totalling £12bn annually. One attraction for Octopus is that GoCardless can process its 30,000 weekly customer refunds far faster than the current provider.
In corporate news, finAPI, the German open banking platform, reported revenue up 23% to €7m in 2024, with a net profit of €129k. That makes it something of an outlier compared with the structurally unprofitable UK open banking sector. FinAPI, now owned by the acquisitive Italian fintech Fabrik, says it processes around €7bn of payments each month and charges roughly €0.14 + 0.5% per transaction.
It’s surprising that anyone still uses cheques, but Germans wrote almost 2m in 2024. That will soon end as cheques are scheduled to disappear by 2027. In France, the government will also stop accepting cheques for tax payments and close its last public processing centre.
Citrini’s arguments are open to debate, particularly its claim that AI agents will move commerce to stablecoins, destroying the card schemes’ business model. I’m not convinced. More likely they will choose Amex for the rewards.
Mastercard deserves credit for its commitment to phasing out PANs in Europe by 2030. I’m less sure the announcement required this video. I hope Peter Schmeichel was well paid.
Where to find me
I’ll be in Berlin 17-19 March for MPE. You’ll find me chairing a roundtable of software/payment convergence and sitting on the payment infrastructure panel. Next, I’ll be moderating a discussion on the business case for payment data at Pay360 in London on 26 March.
There’s been no shortage of commentary about what’s gone wrong at PayPal. Volume growth has stalled, the CEO has been fired and the stock price collapsed following the Q4 results. A weak performance in Germany, where PayPal is the leading eCommerce payment brand, is particularly bad news. Germans normally fund PayPal from a bank transfer not a credit card. This makes Germany a very profitable market for PayPal. The CFO explained: “Our German growth has moderated due to macroeconomic softness, normalization of our long-standing market leadership position and competition from alternative payment methods.”
Worldline’s new CEO indicated he was open to selling further assets. The struggling Paris-based processor will now focus on its European core businesses. The JV with ANZ Bank in Australia would likely be high on the list for sale. He ruled out a merger with Nexi saying “I don’t see for Worldline any priority for further complementary mergers.”
Global Payments completed its acquisition of Worldpay which creates the world’s largest merchant acquirer. The combined business will process a whopping $3.7 trillion from 6m merchant locations. I’m told this is very much a takeover. Global Payments executives are in charge and the Worldpay name will likely disappear.
PagoNxt, Santander’s payment business which operates in Spain, Portugal and Latin America, is now consistently profitable, generating a record €97m operating income in Q4 2025. Payment volume was up 8.5% to €64bn.
PayPoint plc has published its Q3 update including poor results from its UK ISO business. Net revenue fell 3% and volume was down 7%. Management blames “lower than anticipated consumer spending patterns” but “stronger than anticipated competition for SME merchants” might be more accurate.
MPE 2026
If you only make one payments conference this year, make it MPE 2026 in Berlin 17-19 March. I’ve been involved since 2015 and always learn something new. MPE attracts a strong mix of vendors, advisers and merchants and it’s friendlier than most, making it a great place to meet new people in the industry.
This year, I’ll be moderating a panel discussion on payment infrastructure – build vs buy – and hosting a roundtable on software/payment convergence.
Farewell Lyf Pay. The French QR code mobile payment/loyalty wallet looked beautiful but despite blue chip backers including BNP Paribas, Crédit Mutuel, Groupe Casino and Auchan Retail, Lyf Pay never gained much traction and burned through at least €180m. Conclusion: it’s easy to build a wonderful new payment product but much harder to get people to use it.
In corporate news, Corpay is selling PaybyPhone to Lightyear Capital for an undisclosed sum just two years after buying the business from Volkswagen for c.$300m. Paybyphone, a pay-for-parking service operating in UK, France, Switzerland, Germany and North America generates c.$100m annual revenues but its consumer-proposition was looking increasingly peripheral to Corpay’s B2B focus.
Euronet has bought the merchant acquiring business of CrediaBank, the fifth largest bank in Greece. Terms were not disclosed. The sale also includes 20,000 merchants, 2,500 ATMs, issuer processing and a long-term distribution agreement for merchant services through the bank’s branch network. This looks like a very good deal for Euronet. There are clear synergies in merging Credia Bank’s portfolio with ePay, Euronet’s existing Greek business which was based on the acquisition of 200,000 merchants from Piraeus Bank in 2022.
Guavapay latest. The failed London fintech is now in liquidation with an official receiver appointed to salvage what they can for the creditors. Mastercard is owed £17m but my sympathies are with the staff who weren’t paid for September or October. Guavapay’s founder resigned citing “fatigue and health related reasons” but the London-based management has many questions to answer about this mess. To her credit, Laura McCracken – former Chair and then CEO – has answered some of them in response to my post on LinkedIn
In fundraising news, APEXX Global, the London-based payment orchestrator, has raised an additional $10m. APEXX has been winning travel clients recently including et2, Iglu.com and Norse Atlantic. The company’s most recent results show revenue rising 21% in 2024 to £5.7m with an operating loss of £6.9m. Merchants will be interested to note that 38% of sales come from commission payments from acquirers.
Mews, Czech by origin but based in Amsterdam, has raised $300m to support its hotel software suite. Mews processes $20bn annually from 15,000 merchants and the business case is very much payment-powered. In 2024, Mews made 75% of its €209m revenue from transaction-related sales, mostly from commissions on payment processing. Mews is a payment facilitator – handling onboarding and risk – but transactions end up with Adyen and Stripe as acquirers.
Klearly, an Amsterdam-based restaurant payments vendor selling through partnerships with ISVs, has raised an additional €12m. Klearly has a stellar line of up backers including PayPal, and the new money will help fund expansion to Italy.
This has put added pressure on both Wero (a mobile wallet linked to SEPA instant payments) and the digital euro as strategic programmes that will help reduce dependence on Visa, Mastercard and PayPal. The UK evidently doesn’t share these fears and has extended its contract with Mastercard to run Faster Payments, the domestic inter-bank payment network.
Werowatch
Wero, commercialised by the European Payment Initiative (EPI), is certainly making good progress. The helpful werotracker website shows 50 banks in France, Belgium and Germany now supporting live transactions. New banks and PSP announcements are coming thick and fast, most recently Deutsche Bank, Postbank and Mollie. And iDEAL, the ubiquitous Dutch online payment scheme, is already rebranding as wero as you can see in this glossy commercial.
How much has wero cost so far? The FZ blog estimates that EPI’s (a consortium of European banks plus Nexi and Worldline) have put up €670m of capital. EPI made an operating loss of €55m in 2024 and wrote down a further €41m, thought to be related to its investment in Payconiq. There is plenty of money left in the bank.
Wero has signed an agreement with four national schemes – Bancomat (Italy), Bizum (Spain), MB Way (Portugal) and Vipps (Nordics) – to accelerate inter-operability. If successful, this could give Europe a credible cross-border merchant payment capability for the first time. Plans are moving quickly. Co-operation will be based on the creation of an “interoperability hub” operated by a new entity established by this summer. P2P payments are schedule to be available by Christmas with merchant payments coming in 2027.
The digital euro is currently stuck in political ping-pong between the European Parliament and Council. Centre-right and centre-left MEPs can’t agree whether it should only work offline or whether there should be online functionality too. The European Central Bank is planning to pilot the digital euro in 2027 with a full launch in 2029. The ECB estimates it will cost €1.3bn to develop the new currency and will incur c.€320m in annual running costs. The first of its expected 100 staff are being recruited now.
Scheming
Combined Mastercard and Visa volume in Europe (measured in euros not in depreciating dollars) is growing at c.8% compared with a consistent low-teens rate over the last three years.
The slowdown is clear but why? Here’s what Business of Payments readers think.
After the UK left the EU, the card schemes took back control of Interchange paid by British merchants on eCommerce transactions to EU cardholders and increased it from 0.3% to 1.50%. This is now costing British merchants an extra £150-£200m annually. The regulator wants to reintroduce an Interchange cap and just won a court case brought by Mastercard, Visa and Revolut who were trying to keep Interchange high.
Across Europe, the newer, mobile-centric local schemes are certainly performing well as shown by Marcin Mazurek in this report. Poland’s Blik, in which Mastercard has invested, is the star.
Ireland has been lacking a domestic mobile payment scheme but now welcomes Zippay. Built byNexi and based on Pago Bancomat, Zippy will go live on St Patrick’s Day, 17 March. The local banks supporting Zippay hope it can help them fight off Revolut which is particularly strong in Ireland.
Payments and software are now inextricably linked. Evidence of the shift is mounting. Tidemark’s 2025 Vertical and SMB SaaS Benchmark Reportfinds that 87% of vertical software-as-a-service (SaaS) vendors now offer payments, up from previous years, with 31% forcing merchants to take their preferred processor. The median attachment rate has jumped from 23% to 40%.
Lightspeed, the Canadian restaurant software vendor moved quickly to incorporate payments in its standard product. Lightspeed Pay (Adyen behind the scenes) offers integrated payments to customers in UK, France and Germany. Total European revenue (software plus payments) was up 21% in Q4 following the recruitment of 150 sales reps “going city by city.” Lightspeed targets the more complex restaurants and says it consistently makes 40bps net revenue from payment processing.
One reason that Adyen wins software distribution partnerships is the depth of its financial services proposition. Fresha, an ISV serving salons and spas, has chosen Adyen to provide embedded lending to its 140,000 customers worldwide, including UK, Netherlands, Finland and Sweden. This is merchant cash advance. Repayments are taken from the daily card settlement.
There could be trouble ahead. Investors have started worrying that AI will kill the market for vertical software such as retail and restaurant point of sale. Why would merchants rent expensive packaged software from an ISV when AI can write them customised code that does exactly what they need?
A collapse in packaged vertical software would close off this distribution channel for payments. What would replace ISVs? Maybe there’s space for an AI friendly marketplace of payment integrations.
Most people think that new shopping is about speed. Maybe’s it’s the opposite. A Dutch supermarket is having success with “chat lanes” at which shoppers are encouraged to slow down and talk with staff and each other.
Amazon has also given up on palm payments in “response to limited customer adoption.”This isn’t suprising. There are a few use cases for biometric payments – saunas, swimming pools etc – but only where people don’t have their phones with them. Palm payments will never be a mass-market thing.
Agentic Commerce
Agentic commerce is moving faster than a monthly newsletter can cover. The pace of adoption of this new technology has been stunning.
Merchants are faced with some difficult questions. Agentic commerce could drive profitable new business as Adobe shows in its very helpful quarterly AI traffic report. AI initiated shopping converts at higher rates and delivers 30% larger basket size.
But in agentic commerce, retailers no longer compete on who has the best brand and website; they compete on who is most legible, trusted, and valuable to AI agents. This requires them to expose their product catalogue, pricing file and promotions logic which could mean giving away all competitive differentiation and advantage.
This is why Amazon won’t allow agents to access its data. Nor will eBay, which has updated its user agreement to ban “buy-for-me agents, LLM-driven bots, or any end-to-end flow that attempts to place orders without human review”.
The AI companies and their platform partners understand the value they bring and smaller merchants may not be allowed a choice. For example, Shopify has begun asking 4% extra commission on sales made via OpenAI powered checkouts. It’s not yet clear how this would be shared between Shopify and OpenAI.
Product round-up
Stripe has won payments processing at Currys, a British electrical chain with 300 stores, marking its first major omni-channel retail client in Europe. Freedom Pay is providing terminals and transaction routing. The Stripe/Freedom Pay partnership could be the first credible challenger to Adyen’s multi-market omni-channel dominance in Europe.
Polcard (Fiserv’s Polish unit) has equipped 400 reverse vending machines. Tap your card and get instant payment for recycling.
Age verification is always a challenge for self-checkout. Voltox, a Swiss-German start-up, has a slick integration with Android terminals that proves you are old enough to buy alcohol. Here you can see it working on a Castles device.
The European landscape is littered with failed card-linked loyalty vendors such as Bink (UK) and Izicap (France). It’s a business model that promises easy collaboration between merchants and potential customers but has proven almost impossible to make money from.
London-based Krowd may be the first card-linking vendor to reach a sustainable business model. Krowd, which came through the Techstars accelerator, has a restaurant-focused loyalty proposition and powers rewards platforms for American Express and Revolut. Krowd uses the payment account reference (PAR) as the unique identifier that recognises consumers across channels.
SoftPOS
SoftPOS is the technology that allows any Android or iOS device to take card payments. Originally conceived as a way of enabling a long tail of micro-merchants, SoftPOS is proving more exciting for large enterprises. Here are two examples:
Worldline is supplying SoftPOS to Poland’s intercity rail network. The train guards will no longer need to carry a separate payment terminal and can take payments on their existing Zebra handheld computers instead. 1,367 integrated devices are already in use, making 315,000 transactions in the first month of operation.
California-HQ’d Magic Cube, a SoftPOS vendor, has raised $10m for its continued expansion. Verifone is one of the investors. Magic Cube powers Dojo’s SoftPOS product which is claimed to be one of Europe’s largest SoftPOS implementations.
Open banking
UK residents typically make 2.3bn card transactions and 400m direct debits each month. In December 2025, they made just 35m open banking payments. Open banking has a long way to go.
Industry players are pinning their hopes on commercial variable recurring payments(cVRPs) which could replace card-on-file and direct debits. If they did, this would deliver significant volumes. But today, banks cannot charge for open banking payments and so have no incentive to promote them. UK Finance, a trade body, has published its proposals for a pricing structure which, it says, should be ad valorem and include consumer protection.
The report is careful not to recommend an actual number. That would be a red flag for the competition authorities. But reading between the lines, here’s my take:
It appears that the industry wants the regulator to land on a fee of around 20bps for cVRPs. This would be cheaper than debit which has a base cost of c.30 bps including scheme fees. I’d also expect a cap of £1 or £2 per transaction and the ability for large merchants to negotiate bilateral deals with banks. The question now passes to a series of backroom discussion between the UK’s overly complex landscape of regulators and industry bodies.
There’s no indication yet of stablecoins becoming mainstream for merchant payments, either for consumers to make eCommerce purchases or for merchants to be settled transactions.
Visa, which has invested heavily in stablecoin acceptance and settlement, sees applications for treasury management in developing markets but “we don’t see a lot of product market fit in developed digital payment markets like the United States or like the U.K. or Europe for stablecoin payments.” This is backed up by news that Shopify has processed just $600K in USDC (the leading stable coin) since launching in June 2025.
Stablecoin advocates throw out some very big numbers but it’s worth remembering that most of the activity supports trading in and out of crypto currencies. Jeremy Light calculates that just 7% of reported stablecoin volume is used to facilitate transactions. In December 2025, there were 376m transactions at an ATV of $751.
Crypto – stable and unstable – can’t shake off its association with crime, money laundering and tax evasion.
A crypto-exchange based in Caracas called Kontigo has been helping locals evade sanctions. I know. You’re shocked such as thing could happen. You’ll be more shocked at the US fintech establishment folks – Y Combinator, JP Morgan and Stripe – who were up to their eyeballs in this mess. Great reporting from Jason Mikula and Fintech Business Weekly.