On The Wire

Every merchant implementing A2A asks the same question. How long until adoption means something? A consumer electronics retailer with 28 stores and €86 million in revenue has the honest answer, measured every month for three years. Month 1: 2.1%. Month 36: 30.4%. Not a climb. An S-curve, with three phases that behave completely differently. Months 1 to 6, from 2% to 6%: every conversion is manual....

Three years of monthly adoption data from one retailer, including the parts that did not work and a financial verdict that undercuts the usual sales pitch. The context: 28 stores across Germany and Austria, €86 million revenue, 60% in-store and 40% online, €142 average transaction, a tech-savvy customer base, and €1.2 million a year in payment costs against electronics margins of 4-6%. Year one. L...

Payment processing is managed as a cost centre. In a drive-through, that framing costs more than the fees do. An 85-location quick-service chain runs €42 million through drive-through lanes and pays €462,000 in card fees. The obvious project is to negotiate the fee down. The less obvious number: payment takes 38 seconds of a 180-second service time. That is 21% of the lane, spent retrieving a card...

Drive-through is the most time-sensitive commerce environment there is. This episode works through what that actually means for payment, and why the fee is the least interesting number. The economics. Average service time 180-240 seconds per vehicle. Target throughput 15-20 vehicles per hour per lane. Payment takes 25-40 seconds, or 13-22% of the whole interaction. Reduce service time 10% and a si...

Banks that signed merchants to card acceptance in the 1970s still hold those relationships today. That is the uncomfortable precedent for anyone treating A2A as a 2028 decision. The mechanics have not changed. Merchant payment relationships run 5-7 years on average. Annual churn is 10-18%, and merchants who switch almost always do so because something broke, not because they went shopping for a be...

Payment networks are winner-take-most markets. This episode explains why, what the card precedent actually shows, and how long the A2A window realistically stays open. Why payments consolidate to two to four players per market. Merchants will not integrate a dozen payment methods, so they pick the three or four with real customer demand. Customers gravitate to methods that work everywhere rather t...

Interchange fees are not dying. They are being demoted, and the distinction matters for anyone planning around them. The arc is documented. The 1970s: 1-2%, justified as funding card issuance. The 1980s and 1990s: credit goes mainstream, rewards escalate, interchange climbs to 1.5-2.5%. The 2000s: 2-3% at peak, premium cards commanding the highest rates, merchants complaining and paying anyway bec...

This episode is careful about a distinction most coverage blurs: what is documented history, and what is a scenario. The regulatory record is fact. The next twenty years are assumptions, and we name them. The rise, 1970-2015. Card networks form, banks issue, merchants accept, interchange transfers value from merchants to issuers at 1-2%. Rewards programmes escalate through the 1980s and 1990s and...

A grocery chain running 47 stores and €240 million in revenue paid €1.92 million in card fees last year. That is 0.8% of revenue, which sounds survivable until you measure it against profit instead of turnover. Net profit: €2.88 million. Payment fees: €1.92 million. Payment processing consumed 67% of net profit. For every €3 the business earned, roughly €2 went to accepting the payment. Per transa...

Payment costs are treated as overhead - the unavoidable price of doing business. This is the story of a grocery chain that turned them into pricing power, told with the numbers and the failures. The economics that forced the decision. €240 million revenue, €192 million cost of goods, €43.2 million operating expenses, €1.92 million payment processing, €2.88 million net profit. Payment fees were 67%...

Every bank CFO runs the same calculation before rejecting A2A. €7.77 billion in card volume at 0.6% is €47 million in revenue. Shift 30% to A2A at 0.5% and you earn €11.7 million on it, but you lose €14 million of card revenue. Net: minus €2.3 million. The arithmetic is flawless. The conclusion is wrong, because the model counts what you lose and not what you stop losing. It holds everything const...

A European bank with 18,500 merchants and €47 million in card acquiring revenue spent eighteen months not launching A2A, because of one slide showing minus €2.3 million. This episode takes that slide apart and rebuilds it. The five-year model, year by year. Year 0: €7.77 billion of card volume, €47 million revenue at a 0.6% average rate, 12% annual merchant churn, 18,500 merchants. Year 1 is genui...

"Cards cost 2%. A2A costs 0.5%. Therefore A2A is cheaper." That comparison is correct and almost useless. It measures the one line a merchant can see on a statement. Run the whole number on a €10 million European merchant. Direct card fees: €90,000. Then the costs that never reach the processor's invoice. Chargebacks: €8,750 across 250 disputes and 125 hours of staff time. Fraud tooling, 3D Se...

The honest comparison between cards and A2A is not the one on the rate card. This episode runs the total cost of ownership on both, then argues against itself about where cards still belong. Start at transaction level. A €50 debit card payment in Europe: €0.10 interchange, €0.04 scheme fee, €0.20 acquirer markup, €0.10 authorization, €0.05 gateway. €0.39, or 0.78%. The same €50 on credit: €0.55, o...

A three-day festival with 45,000 attendees processes €2.8 million on site. Payment costs reach €177,400. That is 6.3% of volume, and the processing fee is the smallest part of it. The breakdown: €36,400 in card processing, €18,000 renting terminals and connectivity for 72 hours, €14,000 in transactions that failed because the cellular network was saturated by the crowd, €22,000 in cash handling, a...

Retail amortises a card terminal over years. An event rents one for 72 hours and pays again at the next event. That single structural difference is why event payments are the clearest A2A case in commerce, and why the savings are not primarily about fees. The baseline. A three-day festival, 45,000 attendees, €2.8 million across 155,000 transactions at €18 average, 25 payment points. Costs: €36,400...

There is no European A2A adoption rate. There is a 5x spread, and the spread is the strategy. Northern Europe (Netherlands, Sweden, Denmark, Finland, Norway) runs 18-25% merchant acceptance and 8-12% of transaction volume, on low cash usage, high trust in digital banking, and labour costs that make payment efficiency worth chasing. Western Europe (France, UK, Ireland) sits at 12-18% acceptance. Ce...

Most payment strategy decks quote one adoption number for Europe. The number is useless. Adoption varies 5x by geography, 5x by vertical, 5x by age, and inverts by income. The variance is where the resource allocation decisions actually live. Geography. Northern Europe leads at 18-25% merchant acceptance and 8-12% of volume, on low cash usage, established digital banking trust, proactive PSD2 impl...

Card fraud costs European businesses €1.8 billion a year. The industry's response has been to build more layers around the card number. That is the wrong problem. A card number, a CVV and a billing address are all the same kind of secret: something you know. Every one of them can be copied, and once copied, reused. Adding 3D Secure on top cuts online card fraud 40-60%, which is real, but it is...

Most payment security is an attempt to protect a number that has already been copied. The alternative is infrastructure where no reusable credential exists at all. SCA requires two independent factors from different categories. Knowledge is something only the customer knows. Possession is something only they have. Inherence is something they are. Online card payments fail this by design, because t...

A SaaS company with €18 million in recurring revenue reports 8% annual churn. Two points of that are customers who decided to leave. Six points are customers who wanted to stay and were removed by a failed payment. That is €1.08 million a year. Not lost to a competitor, not lost to a bad product. Lost because a piece of plastic reached its expiry date. Break down the cause and it stops looking lik...

Subscription businesses report churn as one number. Inside it are two different problems. One is a product problem: customers evaluated your service and left. The other is plumbing: customers never made a decision at all, and a payment method failed on their behalf. Most teams optimise the first and tolerate the second. Benchmarks: 3-5% involuntary churn for B2B SaaS, 6-10% for B2C, 8-12% for high...

Every market that has regulated interchange has moved the same direction. None have moved back. Europe capped consumer interchange in 2015 at 0.2% debit and 0.3% credit, down from 0.6-0.8% and 1.2-1.6%. For a merchant running €10 million at a 70/30 debit-credit split, interchange fell from €91,000 to €23,000. A 75% cut. Australia started in 2003, taking average interchange from 0.95% to 0.50%, the...

A merchant paying 2% on a €100 card sale is paying three parties. €1.20 of interchange to the customer's bank. €0.20 of assessment to Visa or Mastercard. €0.60 of margin to their own processor. The largest slice goes to a bank the merchant has no relationship with, at a rate the merchant had no part in setting. That last clause is why regulators keep intervening. Every major market that has, w...

Ask an e-commerce merchant what payments cost them and they quote the processing rate. 1.2%. The real number is roughly double that, and most of it never appears on the processor's invoice. A retailer doing €22 million a year pays €264,000 in card processing. They also lose €104,000 to fraud and prevention tooling, €65,000 to false declines where their own filters blocked real customers, €42,0...