An EV charger can be technically available and functioning, and still fail the driver. Richard Hackforth-Jones of Sainsbury’s Smart Charge gave a simple example: a driver arrives at a charger but cannot start charging because the payment terminal does not accept their card. The charger may be functioning exactly as intended, but to the driver, it has failed.
That gap between what the system reports and what the customer experiences is becoming one of the most important challenges in EV charging operations. For years, the industry’s most visible work was securing locations, obtaining grid connections and installing chargers. Once a network is live, a different job begins: making every charging visit dependable.
Across conversations with operators on the AMPECO Ecosystem Stage at ICNC 2026, a consistent picture emerged. The quotes in this piece come from those sessions, all available to watch on demand. The challenge now is knowing when someone cannot charge, understanding why, and intervening before that experience loses a customer.
When a working charger still fails the driver
Looking back on the expansion of Ignitis ON across the Baltics, Eimantas Balta admitted: “The hardware part was the easiest one.”
The harder work was everything around it: uptime, maintenance, billing, systems and customer experience. A small network can survive on manual intervention; at scale, the same issue turns up ten or a hundred times, and that becomes a bottleneck.
For Balta, a reliable network means detecting and resolving downtime before the customer notices.
Balta described a driver arriving at a station to find its cable missing. They might give the location another chance. But if the cable is still missing on a second visit, he expects that the driver is unlikely to return a third time.
The operator may restore availability, but the dashboard cannot show the driver who has already gone elsewhere.
“It’s not just that it’s on, but that it works the first time as well.”
Richard Hackforth-Jones, Sainsbury’s Smart Charge
What your uptime figures aren’t telling you
Uptime remains an essential operational KPI. But operators are questioning how much it tells them about a real charging visit.
At Sainsbury’s Smart Charge, Hackforth-Jones distinguished between the uptime reported under UK regulations, which can exclude certain events outside an operator’s control, and the company’s internal measure. His team also tracks what he called “true uptime,” which accounts for interruptions such as power cuts. To the stranded driver, the reporting distinction means nothing.
A similar challenge came up in Rowald Oosterkamp’s keynote for Velian. He recounted a conversation with another industry executive who questioned whether a 99% uptime target was still the right measure. The more meaningful question, they argued, was charging success: did the person who came to charge actually manage to do so?
Uptime measures infrastructure availability under defined conditions. Charging success rate, in its simplest form, asks what share of charging attempts actually delivered energy. Together, they reveal failures a green status light can hide.
Consider the sequence a driver must complete: find a usable bay, connect the cable, authenticate, pay, initiate a session and receive electricity. A failure at any point can make the visit unsuccessful, even when the charger itself remains online.
For operations leaders, this suggests a different set of questions alongside charge point uptime: How often do charging attempts fail to start? How many sessions end with no energy delivered? Do payment failures cluster at particular sites? Do roaming customers encounter different problems from direct customers? And which of those problems are being resolved before the driver contacts support?
Richard Hackforth-Jones, Smart Charge, on why “works first time” matters more than the uptime figure.
At scale, the exceptions become the real challenge
INDIGO’s experience shows why these questions become urgent at scale. The parking operator has grown its charging business to thousands of charge points across markets. As Global e-Mobility Director Claude Muller put it, managing 50 chargers and managing thousands are fundamentally different tasks.
At 50 charge points, he explained, many tasks can still be managed manually. At a greater scale, the work has to become systematic. For INDIGO, that means building measurable processes around maintenance, escalation, and problems that might once have been handled as isolated incidents.
“It only becomes scalable, really, when exceptions are measurable.”
Claude Muller, Global e-Mobility Director, INDIGO
Muller specifically pointed to zero-kWh sessions, suspended sessions, and disputes with roaming partners. Each can represent a failed experience, a service issue or an unresolved operational task while the charger still reports as online.
Recurring exceptions need an owner, a priority and a resolution process.
A zero-kWh session might prompt an investigation into failed authorization, a charging handshake or a driver abandoning the attempt. A suspended session may need different handling. A roaming dispute may involve systems beyond the charge point itself. The point is to stop exceptions disappearing among apparently healthy stations.
As Oosterkamp said of Velian’s Dutch concessions, where one party answers for operations, “If we have non-working chargers, we do not make any money.” But the effect extends beyond that missed session. Poor experiences can undermine repeat visits to the same network, long after the immediate fault has been fixed.
Claude Muller, INDIGO, on making exceptions measurable at scale.
The problems your management dashboard can’t see
Even a more comprehensive charging dashboard is not the whole answer. Information about customer experience often sits outside the EV charging management platform.
Martin Symes, EV Director at MFG EV Power, called the next challenge “seeing the unseen.” His network can monitor chargers, but he wants a clearer picture of what drivers actually experience, including problems that surface only in customer calls, complaints or other feedback.
His proposed direction is to bring together charger telemetry, back-office records, call-center information, connectivity-provider data and the company’s CRM. That is an ambition he is working toward, not a claim that one tool already solves it all.
Take a site that generates repeated calls about sessions that will not start. The charger may not produce a fault severe enough to trigger the usual alarm. But combine the calls with session records and connectivity information, and a pattern may emerge. That combination makes it easier to decide whether the issue belongs with a payment provider, a connectivity partner, a maintenance team or the operator itself.
Repeated failed starts at a busy site deserve urgent attention, even when overall network availability looks strong.
Symes’s point is that different systems must share useful operational data. Otherwise, the driver connects the dots first, by complaining.
The human side of charging reliability
Better monitoring and clearer escalation can prevent many problems. Drivers still need someone to turn to when a problem persists.
Velizar Velkov of payments provider Worldline told a story from his own experience: almost no battery left, pouring rain, poor phone reception, and a payment failing again and again. The cause was the card stored in his charging app, which had expired. Updating the details solved the problem, but it took him several minutes to identify it.
A payment failure is not a charger fault. From the driver’s perspective, however, the distinction offers little comfort when the battery is almost empty.
Network tokenization can mitigate this problem by keeping stored payment credentials up to date. But it cannot prevent every failure, and drivers still need an understandable way out.
When software reaches its limit, a person has to answer, and they are only as useful as what they can see. Richard Hackforth-Jones described the unglamorous side of Sainsbury’s Smart Charge as “just the payment flows and making sure they go through … and also making sure the data that our call center accesses is true so that they can inform the customer what’s going on.”
A useful handover provides the support agent with the driver’s location, the error, and the steps already attempted, so the driver can get help without starting from scratch.
Reliability begins with getting the basics right
Operators repeatedly returned to the basics. Wilfried Thierry of INDIGO described the essential promise as having a charger available, having it in working condition and offering good value charging. Hackforth-Jones made a related observation: a driver may value finding an available bay more than getting the highest possible power.
For CPOs planning next year’s operating priorities, the most useful benchmark is less glamorous than another rollout milestone: how often a driver who came to charge actually did.
Five habits of operators building more reliable networks
- Measure the completed journey. Put charging success and failed starts alongside uptime and availability.
- Treat exceptions as work. Track zero-kWh sessions, suspended sessions, payment issues and roaming disputes, with clear ownership and escalation.
- Act before the complaint arrives. Use alerts, diagnostics and maintenance processes to shorten the time between detecting a fault and resolving it.
- Connect technical signals with customer signals. Review charger data alongside support contacts, CRM feedback and connectivity information to spot recurring issues.
- Make recovery easy for the driver. Automate routine fixes, but keep a clear route to human help when a situation is urgent or automation reaches its limits.
What CPOs should expect from their charging management platform
None of these habits can rely on heroic manual effort as a network grows. CPOs need operational tools that turn charger events into useful alerts, support remote diagnosis and corrective action, integrate payment and roaming workflows, and make it easier for teams to investigate and resolve exceptions.
AMPECO’s issue detection, automated fault handling, payment terminal integrations and roaming support help provide that foundation. CoOperator, AMPECO’s AI operations agent, helps teams investigate and resolve issues faster, but no platform replaces the operator’s processes or other business systems.
The test is whether the people running the network can use those tools to identify the next customer problem and act on it.
The second job is what earns the third visit
Deployment makes a charging network visible. Day-to-day operations determine whether it becomes dependable.
A customer does not see a CPMS, an alert rule or a dispute-resolution workflow. They see a usable bay, a payment that goes through, electricity reaching the vehicle and a solution when something goes wrong. When those things work, there is little to remember. When they fail twice, as Balta warned, there may be no third visit.
That is the second job of a CPO: making the entire charging experience dependable enough that drivers choose to come back.
