AFIR settled an argument the EV charging industry had been having with itself for years. Public charging can no longer hide behind your app. Above a certain power class, a new public charger must accept ad-hoc payment — a driver pays for a single charge with no app, no account, no prior relationship. Walk up, pay, charge, leave. The regulation doesn’t suggest it. It mandates it.
That sounds like a small compliance checkbox. It isn’t. It quietly retires the business model a lot of networks were built on, and it forces a decision most operators have been deferring: when a stranger pulls up to your charger, how exactly do they pay?
What “ad-hoc” actually means
Ad-hoc is the regulator’s word for charging without a contract. The driver and the operator have no standing agreement — no membership, no stored card, no loyalty tier, no onboarding funnel. It’s the relationship you have with a parking meter or a vending machine: a one-time, anonymous purchase.
This matters because public charging is, increasingly, a network of strangers. The driver at your charger might be a tourist, a fleet vehicle, a rental, someone passing through on a road trip, or a local who bought their first EV last week. The regulator looked at that reality and decided the charger has to serve all of them — not just the subset who already downloaded the right app.
AFIR doesn’t leave the mechanism entirely open, either. For higher-power points it specifically requires contactless card readers — a physical terminal that takes a tap. Below that threshold it’s more permissive: a card-based device that generates a payment link or a dynamic code can satisfy the obligation, and mobile-wallet acceptance counts (AFIR Q&A). So the regulation already leans toward the card as the instrument every driver carries. What it won’t accept is a charger that only serves people who came pre-enrolled.
After AFIR, ad-hoc isn’t a feature you bolt on for completeness. It’s the default mode of a public charger. Everything else is opt-in on top.
The hidden tax of “download our app”
The instinctive answer — the one most networks reached for first — is the app. Put a code on the charger, the driver scans it, installs your app, creates an account, adds a card, and charges. Clean, branded, full of upsell opportunities.
It’s also where conversion goes to die.
Think about the actual moment. A driver is standing in a car park they’ve never been to, in front of a charger brand they’ve never seen, often in weather, sometimes in a hurry, occasionally abroad. You are asking them to:
- find the code and scan it,
- leave for an app store and wait for a download,
- open the app and accept terms,
- register an account and verify an email or phone,
- enter card details,
- and only then start charging.
Every one of those steps is a place to abandon. For a driver who will use your network exactly once, the entire account you’re forcing them to create is pure overhead — value to you, friction to them.
The occasional and visiting driver is precisely the driver AFIR is protecting, and precisely the driver the app flow serves worst. A frequent user might tolerate the install once and amortize it across a hundred sessions. A first-timer or a tourist won’t. They’ll drive to the next charger — or worse, conclude that public charging is a hassle, which is the exact perception the regulation exists to prevent.
“Download our app” is a tax. The driver pays it in time and friction; the operator pays it in lost sessions.
The credential every driver already has
QR codes don’t escape this. A static code can be stickered over or swapped — a spoofing vector that points the driver at a lookalike payment page. A dynamic code is more secure, but scanning either one almost always lands the driver in an app or a web payment form: the same friction, just reached by camera instead of search.
Plug & Charge is the most elegant idea in the room: the cable itself authenticates the car, so the driver does nothing but plug in. When it works, it’s the best experience there is. But it needs the vehicle to support it, the car to be provisioned, and the driver to hold an eMSP contract the operator can roam against. That’s a chain of preconditions. Rollout is partial, plenty of cars on the road can’t do it at all, and by design it ties the driver to a contract. Plug & Charge is a wonderful option. It can’t be the floor, because it can’t cover every car or every driver.
Which leaves the one credential that already covers every driver: a payment card.
Nobody pulls up to a public charger without a way to pay — a card in a wallet, or the same card living inside a phone. No install, no account, no provisioning, no contract. The driver already knows how to use it, because they use it everywhere else. A card terminal on the charger asks the driver to do exactly what they do at a fuel pump, a toll booth, or a coffee counter: present a card, approve an amount, done.
This is the spirit of what AFIR is reaching for. Read the regulation charitably and it looks like an attempt to make paying for a charge as ordinary as paying for anything else. A terminal that accepts a contactless tap — built on the same EMV contactless handshake drivers use everywhere — is the most literal, lowest-friction way to meet that. It works for the local and the tourist, the first-timer and the veteran, the EV owner and the fleet card. There is no segment of driver it excludes.
Card payment is a default, not a fallback
Here’s the part operators tend to get backwards. They treat the terminal as the fallback — the thing you offer for people who refuse the app. App first, card if they insist.
Invert that. The card terminal should be the default path, available to every driver, with the app as the value-add on top for the ones who want loyalty, history, and saved preferences. AFIR doesn’t ask for a grudging fallback; it asks for ad-hoc payment to be a real, first-class way to charge — and there’s an open standard built for exactly that: the OCPI DirectPayment module defines how a terminal at the charge point settles a one-off, ad-hoc tap. A fallback you bury behind the app is a fallback that fails the exact drivers the regulation cares about.
Make tapping a card the front door. Let the driver who wants more walk further into the house.
Card-present is better economics, too
There’s a reason this isn’t only a UX argument. When a driver taps a physical card on a physical terminal at the charger, that’s a card-present transaction. The card is there. The chip or contactless handshake happens locally, under EMV. Card-present generally carries the lower risk profile — and typically the better economics — that the payments world reserves for in-person sales.
Push the same driver through a web form or an in-app checkout and you’ve made it card-not-present. The card isn’t physically present; the rails treat it as riskier, and the markups that tend to come with that risk eat into the session. For charging — where plenty of sessions are small — any fixed-cost component of a card-not-present markup bites proportionally harder the smaller the charge. The web flow you reached for to avoid the terminal quietly makes every small charge worse.
So the terminal tends to win twice. It’s the lowest-friction experience for the driver, and it’s card-present economics for the operator. The app-and-web path loses on both: more friction and worse rates.
It’s worth being precise about who does what here. A card terminal welded into the charger is not the only way to land a card-present tap — and often not the best one for the CPO. Chargers are built by people who are exceptional at building chargers, and the charger talks to its back-office over OCPP; that competence is exactly why payment acceptance, fiscal receipting, and card-data handling don’t belong fused into the hardware. Keep the payment layer neutral and the terminal interoperable, and the CPO keeps the card-data scope minimized and the freedom to choose acquirers, back-offices, and hardware without re-engineering the charger.
The ordinary purchase
Strip away the industry vocabulary and the goal is mundane: make charging an EV feel like buying anything else. Park. Walk to a charger you’ve never seen. Enter the amount you want to spend. Present your card — that amount is pre-authorized. Charge. Get progress updates as it runs, a fiscal receipt at the end, and the freedom to stop early by unplugging or pressing a button.
No download. No account. No contract. No relationship you didn’t ask for.
That flow doesn’t need the driver to know anything about your network, your CSMS, your acquirer, or your charger vendor. It needs them to know how to use a card — which they already do.
AFIR made ad-hoc payment mandatory and pointed squarely at the card to deliver it. Every other method is something you offer in addition: Plug & Charge for the cars that support it, the app for the drivers who want a relationship, a code for the corners a reader can’t reach. The card terminal is the one method that excludes no driver, adds the least friction, and bills as the card-present sale it actually is.
So when a stranger pulls up to your charger, there’s only one honest answer to how they pay: the same way they pay for everything else.