Matthias SpeicherIn short
Charging infrastructure is not a product business. It is a fixed-cost-driven infrastructure system. Viability is not decided by price but by the volume of returning sessions. Margin is a result, not a lever. Value comes from throughput, not from stock. And trust is the only structural lever that at the same time reduces fixed costs, increases volume and defuses price wars.
In Part 2 I described how users build and lose trust, and why return visits don’t come from conscious decisions but from the absence of friction. In this part I shift perspective. What does all of that mean for operators? And why is the seemingly obvious answer, lower prices, the wrong question to be asking?
The wrong question
From the operator’s perspective, charging infrastructure is often reduced to one apparently simple question: are we cheap enough? That question is measurable, comparable, communicable. Strategically, it is still irrelevant.
Economic viability is not decided by price. It is decided by the volume of returning charging sessions. The central bottleneck is not the kilowatt-hour. It is utilisation.
The reason is straightforward. Charging infrastructure is not a product business. It is a capacity-bound infrastructure system with an extremely high fixed cost base. The majority of costs arise before the first session ever takes place. Sites need to be developed, grid connections built, hardware installed. Backend systems, maintenance and incident response run continuously, regardless of whether anyone is charging or not.
The overwhelming share of costs is fixed. The only degree of freedom is throughput.
How brutally fixed those costs are
One example makes the mechanics visible. For large consumers, the capacity charge is not based on average consumption. It is based on the highest power draw in a single 15-minute interval of the year. That one quarter of an hour sets the cost band for the following twelve months.
A site can run steadily at 280 kW all year. Draw 300 kW once, and it moves up a band, and that band applies for twelve months from that moment. Some grid operators work in blocks, so the bill jumps.
Taking the published tariff sheet of SH Netz, medium voltage, under 2’500 annual utilisation hours, that costs roughly 10’400 euro per year for 300 kW. To reach the much cheaper band above 2’500 full-load hours, the site would need around 28 percent utilisation of the connection. This is why load management is not an efficiency exercise for an operator. It is fixed cost control. Peak shaving and battery storage keep a site in the cheaper band.
The pattern holds for the whole cost structure. What is fixed is fixed before anyone plugs in.
Why margin is not a scaling lever
In classical business models, margin is the central lever for economic viability. In charging infrastructure, that logic breaks down.
Willingness to pay is structurally limited. By high price sensitivity, low differentiation, high comparability and regulatory pressure. At the same time, a higher price changes nothing about the cost structure. It does not reduce fixed costs, does not increase utilisation, does not improve scalability.
Margin is a result, not a lever. Whoever tries to force economic viability through price shifts demand in the short term. Long term, they lose volume. Without volume, there is no fixed cost degression.
And why does price still sit at the centre of so many strategies? Because it is visible. Reliability cannot be assessed upfront. Session completion cannot be compared. Expectation consistency only develops through experience. Site quality only reveals itself on arrival. What remains is the tariff. That is why it gets overrated.
The calculation you have to turn around
The usual question is: what price do we need for this site to carry itself? I turn it around. How high may the fixed costs be at all, for a site to carry itself at a given utilisation and a given price?
Take a 300 kW station with two charge points sharing the power. Manufacturers quote 95 to 97 percent technical availability, so roughly 350 operating days. In theory that is about 2.5 million kilowatt hours a year. That number is fantasy. It assumes the charger delivers full power every hour of the year.
Real utilisation values per charge point sit between 3 and 20 percent. A Monta whitepaper gives European averages of 150 to 175 kWh per day per charge point, good sites around 225, top performers around 300, and Norwegian exceptions up to 500. For variable costs, meaning electricity, grid fees and levies, roughly 0.15 euro per kilowatt hour is realistic.
Now the numbers work. At a selling price of 49 cents, the contribution margin is 34 cents per kilowatt hour. At 5 percent utilisation, roughly 124’000 kWh a year, the site can carry a maximum of about 42’000 euro in fixed costs. At 10 percent it is 84’000 euro. At 20 percent, a good site, around 169’000 euro. Without a single cent of profit.
Halve the price to 29 cents and the contribution margin drops to 14 cents. The 5 percent site now carries 17’000 euro. The capacity charge alone has already eaten 10’400 of that.
These figures are a model, not the calculation of a real operator. No company publishes its ground rents, discounts and strategic terms. But the principle holds regardless of the inputs. Viability is not determined by price. It is determined by utilisation. A high price per kilowatt hour is rarely a sign of expensive electricity. It is almost always a sign of poor utilisation.
Not stock creates value, but throughput
Many operators measure success by the number of sites, the number of chargers, installed capacity, coverage. That is understandable, because those numbers are visible, communicable and discussable with investors. Economically, they are still not what matters.
The decisive question is not “how many chargers do we operate?” It is: how many sessions take place per site per day?
A small site with high return rates is economically stronger than a large charging park with low utilisation. And whoever thinks the calculation through to the end recognises: expansion does not scale. Price does not scale. Margin does not scale. Only volume scales.
Additional utilisation is for a CPO a survival necessity, but precisely because of that also the attack surface for large roaming partners. Whoever depends on EMP visibility to fill their sites is trading pricing power for utilisation. That is not a criticism. It is the structural core conflict in the market.
The site as entry ticket
And let’s be honest: the site determines whether utilisation is even possible. A site at a heavily trafficked motorway junction or a high-frequency retail location has the potential for strong utilisation. The site is expensive, but it is the entry ticket to the market.
A mediocre site with surprisingly high utilisation is always economically stronger than an expensive premium site sitting empty because break-even is never reached. That sounds obvious. In practice it gets decided the wrong way around on a regular basis, because visibility and prestige are easier to communicate than throughput metrics.
Trust as a structural operating lever
This is where the circle closes back to the user perspective from Part 2. From the operator’s side, the question is: am I consistently well-utilised? From the user’s side: will I come back here? Both questions describe the same system.
Utilisation cannot be optimised directly. You can lower prices, run marketing, build new sites, develop apps. The only way to increase utilisation directly is to drive there yourself and plug in. Everything else works only indirectly, through trust, through return visits, through usage.
Trust operates on three levels that are directly measurable economically. As a barrier to entry, because in mature markets a missing trust history makes market access harder for new operators. As a price driver, because users pay more for predictability than for kilowatt-hours. And as a scale driver, because growth comes from higher usage at existing sites, not from expansion alone.
Trust replaces marketing, subsidies and in many cases a hard price war. Not as image. As market structure.
The consequence for growth decisions
Growth without this mechanism is blind. More sites increase fixed costs. More chargers increase idle costs. More coverage increases complexity. None of that automatically increases trust.
Growth without differentiation dilutes quality. Diluted quality destroys return visits. Without return visits, no volume. Without volume, no fixed cost degression. This is not theory. It is the mechanics of the market.
In the next part I look at the market as a whole: what separates a mature charging market from an immature one, and what can be derived from that for the DACH region.