Is Your In-House Platform Still the Right Choice?

Executive summary

The business decision you never have time to revisit

Running an EV charging business today means contending with change from every direction at once: shifting policies, regulatory and compliance requirements that pile up state by state, advancing technology, and competitors entering markets you already serve. When the urgent list refills every week, deliberately reevaluating a decision that still works never quite makes it to the top.

For many operators, the EV charging management platform is that decision. Sessions complete, payments clear, and the network stays online. So the question of whether it is still the right software strategy for the business gets deferred, not by choice, but by default.

This guide makes the opposite case: taking the time to rethink your software strategy may be one of the highest-leverage moves still fully in your control. You cannot control policy, EV adoption, or what your competitors build. You can control how your business operates, and the platform beneath it increasingly determines your efficiency, speed to market, customer experience, and the opportunities you can pursue.

The stakes of that decision have grown

When many operators made their software decisions, the platform’s job was clear: keep chargers online, complete sessions, and clear payments. Coverage was scarce, reliability was the differentiator, and expansion was a reasonable proxy for progress.

None of that holds in the same way today. Networks are growing, but utilization is not rising with them, and reliability is becoming table stakes rather than an advantage. The reshaping is not hypothetical: in the space of three years, and for different reasons, Shell bought and then sold a charging network, Enel X Way closed its North American business, Tesla opened the Supercharger network to drivers from competing automakers, and IONNA took over operation of Circle K’s US charging portfolio. Capital has become more selective, and who owns and operates charging in North America is still being decided.

The question facing operators has shifted from “where can we build?” to “what can we do with the network we have?”

That changes what the software underneath has to do. The platform is no longer judged by whether it keeps the network running. It is judged by how well it helps the business perform and adapt: how quickly you can change pricing, launch a new commercial model, serve a new segment, integrate a partner, enter a new state, or pursue an acquisition. Those are commercial questions, and the software underneath the operation increasingly determines how easily you can answer them.

If you built your own platform, this guide is for you

Many operators built their EV charging management platform in-house because the software they needed did not exist to buy. In many cases, that was the right decision. This guide does not argue otherwise. It argues that the decision deserves to be judged against the market you operate in today, not the one in which the platform was originally built.

But two things have moved since that decision was made. Commercial platforms have caught up considerably, narrowing the gap that justified building. And the maintenance surface has grown: protocols, payment requirements, metering rules, state programs, hardware integrations, and security standards all have to be kept current, whether or not that work advances the business.

A standard and the evidence to test it

This guide gives you a method, not a verdict. The standard is three tests that apply to any platform, in-house or commercial:

The evidence is four numbers you can pull from your own organization:

Engineering allocation

Of the engineering capacity you used last quarter, how much went to work that moves the business forward, such as pricing changes, customer requests, new segments, and partner integrations, and how much went to keeping pace with the market: protocol versions, payment requirements, metering rules, state program obligations?

The first kind of work creates an advantage. The second only keeps you eligible to operate. If nobody can put a number on that split, that is usually an answer in itself.

Commercial speed

How long did the last commercially important idea take to launch?

Roadmap deferral

How many large customer requests have you turned down, or pushed back more than once, because the team was tied up with OCPP, OCPI, or downtime work? If you can name one, you already know the pain.

Strategic flexibility

If you acquired a network next quarter, how quickly could you move it onto your software stack? In a consolidation wave, investors are looking for teams that can prove they can grow organically and through acquisition.

Together, these numbers make visible the costs that rarely appear in a budget: opportunities delayed, constrained, or abandoned because the platform could not support them.

That last question is the sharpest test in the guide. An acquisition asks the platform to support a business it was never designed for, and that is where in-house platforms usually struggle. Whether or not a deal is on your roadmap, the answer tells you how much strategic freedom your software is giving you, and how much it is subtly taking away. Chapter 7 takes this up in detail.

A note on our perspective

AMPECO is a software company. We provide a white-label EV charging management platform for 200+ network operators across 70+ markets. Our customers operate under their own brand and retain ownership and control of their EV driver data and customer relationships. We do not manufacture chargers, own sites, or operate a charging network anywhere in the world, so we are never your competitor for sites or drivers.

Working with that many operators, across different hardware combinations, commercial models, and regulatory environments, gives us a deep perspective on the challenges network operators face, and a clear view of the role a modern EV charging management platform can play in meeting them.

It also gives us a point of view, and you should weigh it as such, but the method in this guide stands independently of that point of view. Every test and every number can be applied to any platform, including ours, and you can run all of them without even speaking to us.

How to use this guide

Read the chapters in order or start with the one that matches the pressure you feel most. Pull the four numbers from your own organization. Score what you run today against the three tests, using the In-House CPMS Benchmark and the CPMS Evaluation Criteria Worksheet.

Then, based on evidence, decide whether your current software strategy is still the right one.

The status quo may still be the right decision, but make sure it is an informed decision, rather than the default.

A note on terminology: There is no single term used across the North American market for the software that manages an EV charging network. Throughout this guide, we use the terms EV charging management platform, charging management system, platform, and software interchangeably to refer to the core software layer.

How the market has changed since you built your EV charging management platform

When you decided to build your platform in-house, the market rewarded a different set of capabilities. Coverage was scarce, reliability was the crisis, and the platform’s core job was clear: keep chargers online, complete sessions, and clear payments. That was enough for the market at the time, but it is no longer enough on its own.

Today, operators are competing in a denser, more selective market where utilization, customer experience, commercial flexibility, and operating efficiency matter alongside reliability. The platform is no longer judged only by whether it keeps the network running, but by how much it helps the business perform and adapt. Here is what changed.

Expansion stopped being enough on its own

For years, expansion was progress, or close enough: more sites meant better coverage, a larger addressable market, and more opportunities to capture charging demand. That relationship has become less straightforward as networks have expanded and competition has increased, and the recent numbers clearly show the change.

US public fast charging, Q2 2026. Sources: Paren, EVgo Q2 2026 reporting.

More sessions are arriving on a network that is also getting larger, so demand is broadly keeping pace with capacity rather than outstripping it. EVgo’s own reporting shows a similar pattern at the operator level: network throughput increased 13% to 99 GWh, while average daily throughput per stall slipped to 276 kWh from 281 a year earlier.

Adding chargers can still grow the business. What has changed is that expansion alone no longer tells you whether the business is getting stronger. The more useful signals now are those produced by the network you already have: utilization, energy throughput, site-level economics, and the operating efficiency behind them.

The competitive map changed without requiring new sites

The physical footprint is not the only thing that changed. As of Q2 2026, Tesla accounted for just under half of the active US public DC fast-charging port base, while its Supercharger network was opening to a growing list of other automakers, and Paren reported that NACS represented 22.9% of new connectors deployed by non-Tesla networks in Q2 2026, up from around 10% a year earlier.

Read that as a competitive event rather than only a driver-convenience story. A charging footprint once exclusive to Tesla drivers became available to customers of other automakers, changing the competitive map without those networks building those sites themselves.

The same period brought consolidation and strategic retrenchment

Different circumstances, but together they describe a market in which ownership, operating models, and competitive positions are still being reshaped. Meanwhile, competition at the site level has intensified as more networks enter established markets. The structural pressures behind that shakeout have been building for longer than the exits suggest.

Reliability stopped being enough to differentiate

For much of the last decade, many of the industry’s hardest problems were physical. Chargers broke down, sessions dropped, and drivers arrived at stations only to find they could not charge. Investment went where the pain was most visible: higher peak output, power sharing across dispensers, better hardware, and better reliability.

That was the correct priority, and the industry has made meaningful progress. Paren’s Q2 2026 data shows stronger reliability across much of the US fast-charging network, driven in part by more reliable new ports coming online. Its conclusion is the important one: reliability is becoming table stakes. Poor reliability can still damage utilization, but good reliability alone does not guarantee it.

That shift has another consequence. Because reliability was the crisis, customer experience was often reduced to one question: did the charger work? That left less competitive pressure around everything else a driver goes through: how they find you, how they authenticate, how they pay, whether the pricing makes sense, and whether anything about the visit connects to the business they are parked in front of. As a reliable session becomes the baseline, everything around it matters more.

This leaves a question that network operators have not always had to answer

A driver in a competitive metro has options, and a charger that works is increasingly the baseline rather than the reason to come back.

If a driver has several networks to choose from, why yours?

The competitive question is no longer only about where we can build. It is increasingly about what we can do with the network we have.

For many operators, location and price still do much of the work, but neither is an easy advantage to defend indefinitely. Location advantages weaken as competitors build nearby, and competing on price for a capital-intensive asset is a difficult way to make a living.

There is a third source of advantage that much of the market has yet to develop fully: differentiation. Deep pockets and strong sites are both scarce and both contested. Much of the market still competes primarily on proximity and price, which means there is considerable ground for operators that can create a more distinctive proposition. Unlike a site, it is not something a competitor can simply buy next door.

Why this is a useful moment to reassess

The policy environment has compounded these pressures. Federal support has become less predictable, incentives have shifted, and EV adoption has grown more slowly than many earlier projections anticipated. For network operators, the practical effect is less room to assume that weak utilization will correct itself simply as more volume arrives. Increasingly, operators have to create that improvement themselves.

That cuts both ways. A slower buildout also keeps the competitive ground open longer than it might in a boom. Network operators that use this period to strengthen how they run, improving utilization, customer experience, commercial flexibility, and operating discipline, can enter the next growth cycle in a stronger position than those that spend the same period waiting for conditions to improve.

The role of the platform has changed

The operational fundamentals still matter, but they no longer define the full scope of the job

• What pricing structures can you run, and how quickly can you change them?
• Can you build a loyalty program with a retailer or serve a fleet without a development project each time?
• Can the business introduce a new proposition without every change becoming an engineering exercise?
• Do you know which sites are profitable and why?
• If the business changes shape, can the platform change with it?

Those are commercial questions, and the platform underneath the operation increasingly determines how easily you can answer them. Sites, capital, and execution still determine a great deal in the EV charging business. Software has not replaced any of them, but it has moved from a supporting system toward a layer where a meaningful share of differentiation, flexibility, and commercial speed now lives.

That is why the criteria used to judge it deserve another look. Before we can define that standard, there is a more fundamental question: what do you actually need to own to retain control of your EV charging business?

What should you own in your EV charging business?

If you built your own EV charging management platform, you made a deliberate choice to own one of the most important layers of your charging business. At the time, that may have been the only way to get the control, flexibility, or functionality you needed.

EV charging platform vendors often span more than one part of the stack. They may sell the hardware, provide the software, operate a driver network, or do several of those things at once. That can make the alternative to building in-house look binary: either own the software yourself or hand more of the business to a vendor.

But there is a third option: use a specialist EV charging management platform for the foundational software without giving up control of the business. You retain control of your hardware, sites, brand, pricing, data, customer relationships, and commercial strategy, and can continue building the capabilities that make your business different.

Every EV charging business runs on three layers

Business layer

Determines how the network competes. It covers brand, customer proposition, commercial model, partnerships, and pricing strategy.

Software layer

Runs the charging network. It manages chargers, sessions, pricing, payments, roaming, energy, data, and the driver experience.

Physical layer

Gets charging into the ground. It covers site development, utility connection, construction, electrical infrastructure, and the chargers themselves.

For an operator already running an in-house EV charging management platform, this raises a more specific question. It is not whether the software layer matters; you already know that it does. It is how much of that layer you need to own yourself in order to retain control of the business and create an advantage.

Owning the business does not require owning every line of code

A turnkey provider’s own network does not make them a bad partner, and plenty of operators work with them successfully. But it is a structural feature of the relationship rather than a matter of goodwill, and it is worth understanding before you build a business on top of it.

For operators accustomed to building in-house, the specialist model matters. Moving the core platform to a specialist provider can shift engineering capacity away from maintaining the platform and toward the products, integrations, and customer experiences that create competitive advantage.

Where AMPECO sits

We focus solely on the software layer. We do not manufacture chargers, own sites, or operate a charging network anywhere in the world. That has four practical implications:

We do not operate a competing network

We are not bidding against you for sites or competing for your drivers. Our role is to support the business models you choose to run, rather than balance them against the interests of a network we operate ourselves.

Our investment is focused on the software

We do not have a hardware business or charging network competing for that focus. Our R&D is directed at developing and maintaining the platform.

Your business, your brand

White-label throughout, so drivers interact with your app, your name, and the experience you choose to create. You retain the customer relationship and control of your EV driver data. It goes further than a logo and a color scheme, which is where most white-label arrangements stop.

Your differentiation is yours to build

API-first, so your team can extend it, connect it to the systems you already use, and build proprietary capabilities without waiting for every idea to be added to our roadmap.

The point is not that AMPECO should define what makes your business different. The platform should give your team the foundations to build that difference themselves.

Where does ownership create an advantage?

The decision is often framed as a choice between control and dependency: build the platform yourself and keep control, or buy it and give up some of that control. But controlling the business and owning all of the software underneath it are not the same thing.

Some parts of your software may genuinely create a competitive advantage and deserve to remain proprietary. Others may be necessary to operate but offer little differentiation, even though they consume substantial engineering capacity to build and maintain. The boundary between those two categories can move over time. Capabilities that once had to be built in-house may now be available commercially, while the requirements placed on the platform continue to grow.

That does not mean building your own platform was the wrong decision, or that replacing it is now the right one. It means the original decision is worth reassessing against what your business needs today. Before you can make that comparison, you need a clear standard for what the software layer should actually be able to do.

What a modern EV charging management platform should enable

Once you separate control of the business from ownership of the software, the question becomes simpler: what should an EV charging management platform enable the business to do today, as it grows, and when requirements change? The standard should be the same whether you built it in-house or use a software provider.

Three tests your platform should pass

A useful way to evaluate any platform, in-house or commercial, is to ask what it lets you do across three time horizons.

Start with the fundamentals. Set aside growth, new segments, and future business models for a moment. Can the platform run the network you own today, reliably and at scale, without constant engineering intervention?

Sessions need to start and complete successfully. Faults need to surface quickly, with the tools to diagnose and resolve them. Payments need to work across the methods you support, with accurate settlement, reconciliation, refunds, and disputes. Payment requirements also need to be met where they apply, including statewide requirements in California and Washington, as well as those attached to publicly funded sites in North America.

This is the baseline. If the platform cannot reliably run the business you already have, the rest of the evaluation does not matter.

This is where the differences between platforms become more meaningful. Much of a charging business’s commercial flexibility depends on how quickly the software can adapt, and that comes down to three questions.

Charging data becomes more valuable when it can move into the CRM, ERP, energy management systems and other tools already running the business: connecting charging activity to fleet billing, customer accounts, site economics, energy costs, or partner reporting rather than leaving those workflows isolated inside the platform. The same principle applies to APIs and customer experience. The platform does not need to anticipate every product you will build, but it should give your team enough access to create new integrations, interfaces, workflows and pricing models without changing the core platform each time.

Advantage is not only about what you can build for customers; it also shows up in what you can protect. In North America, a recent example in New Hampshire and regional tariff analysis covering the Kansas City market show how demand charges can make the timing of high-power charging materially important to site economics. At an otherwise lightly loaded site, a handful of high-power sessions can set a peak that affects the month’s demand charge. Dynamic load management, pricing, and demand-response integrations can therefore protect margin on sites you already operate.

And the data should help answer commercial questions, not simply operational ones. Which sites are profitable? Which customers are worth retaining? Did a pricing change improve margin? Is weak utilization a demand problem or a location problem?

The goal is not more functionality or more reporting. It is giving the business more ways to act.

The hardest requirements to plan for are often the ones you cannot see coming. Your hardware mix will change. New manufacturers, charger models and connector standards will enter the network. A platform that works with OCPP-compliant chargers, and keeps pace with newer versions of the standard, gives the business more freedom to adopt new hardware without turning every change into a bespoke integration project.

Your business may change too. An acquisition could bring chargers you do not support, commercial models you have never offered or billing requirements your platform was never designed to handle. And technology will keep moving: smarter energy management, AI-assisted operations, new cybersecurity requirements and services that do not yet exist.

You cannot predict every requirement that will matter five years from now. You can assess whether new ones are absorbed as part of normal evolution, or whether each one starts another rebuild.

Breadth matters. So does depth

The goal is to strike the right balance: enough breadth to handle common requirements without constant custom development, and enough depth to support the complexity that comes with growth and build the things that make your business different.

There is a trade-off here. More capability can create more complexity for the operator if it is difficult to configure, understand, or use. A modern platform should absorb as much of that complexity as possible rather than simply transferring it from the engineering team to the operations team. It needs to handle the complexity that should be standard, while giving your team the flexibility to build what comes next.

Applying the standard

None of this argues for a particular platform, including ours. It argues for a standard. Apply the three tests to whatever you run today, in-house or commercial, and keep them as the frame for the rest of the guide. Once you have that standard, the next question is whether continuing to own the platform is still the best way to meet it.

Score the standard with the Platform Evaluation Criteria Worksheet

A good decision can stop being the right decision

When you reconsider your software strategy, the instinct is often to focus on the platform itself: what needs fixing, upgrading, or replacing? But if you run an in-house EV charging management platform, there is a more fundamental question: is owning the platform still creating an advantage for your business?

When ownership creates an advantage

Ownership has real value when it gives the business capabilities it cannot get elsewhere, lets the team move faster, or supports something genuinely proprietary. In those cases, maintaining the platform yourself may still be the right model.

The question becomes harder when the benefits of ownership stay roughly the same, while the work required to sustain them keeps growing. That is when control on paper can mean less control in practice. That is increasingly the tension for established in-house operators: how much of the development burden can be reduced without giving up the competitive advantages that made ownership valuable in the first place?

What changed in the ownership equation

Commercial platforms caught up considerably

Many operators built in-house because the functionality they needed was not available to buy. That gap has narrowed substantially. The question is no longer what the market couldn’t provide then, but what genuinely requires proprietary development now.

The maintenance surface grew

Protocol versions, payment requirements, metering certification, security standards, program reporting, hardware integrations and roaming relationships all have to keep moving with the market. Much of it is necessary, but these items rarely create a competitive advantage.

The trade-off changes over time

The decision to build may have been made years ago, but the work of owning the platform continues. None of that makes ownership inherently inefficient. It does mean the economics and organizational burden can look very different five or ten years after the original decision. The useful test is whether the benefits of ownership still outweigh the burden and cost of maintaining it.

For many established operators, the first of Chapter 3’s tests may be the easiest to pass. The signals below are more useful for understanding where problems with the second and third can begin to appear.

Maintenance is consuming a growing share of engineering capacity

A platform should create operating leverage as your business scales. If maintaining it requires steadily more people just to keep pace with your network and the market, the economics deserve another look.

Control is no longer producing speed

Owning the roadmap sounds like maximum flexibility. But if pricing changes, integrations, new propositions or new segments repeatedly wait behind platform work, the practical result may be less flexibility, not more.

The architecture increasingly reflects the business you used to be

A system designed around one hardware mix, customer type or commercial model may run that operation extremely well. The constraint appears when the business moves beyond what the original architecture was designed to support.

Core software work is competing with differentiation

Time spent maintaining foundational charging functionality cannot also be spent on the customer experience, partnerships, commercial products and proprietary capabilities that set you apart.

A commercial platform is not automatically the cheaper option. It has its own costs, dependencies, and trade-offs. What changes is the basis of comparison. The economics of build versus buy include not only direct software costs, but also maintenance, time to market, engineering capacity, and the value of what that capacity could produce elsewhere. Dependency matters too, and Chapter 10 provides a framework for evaluating it.

The question is not simply what the platform costs to run. It is what continuing to own it consumes, delays, or prevents the business from doing.

Those costs are harder to see, but they may be the ones that matter most.

The hidden cost of owning your EV charging software

When operators calculate the cost of keeping an EV charging management platform in-house, they tend to count what they can see: engineering salaries, infrastructure, compliance and maintenance. That is only part of the cost. The harder costs are what maintaining the platform consumes: engineering capacity that cannot go elsewhere, capabilities that arrive later than they should, knowledge concentrated in a few people, and opportunities the business may never get far enough to pursue.

Those costs rarely appear in a budget. Over time, however, they can become just as consequential as the ones that do. Five hidden costs are worth examining.

Some engineering work creates an advantage. Some simply keeps you eligible to operate. For an in-house platform team, there is a steady stream of work in the second category: protocol changes, payment recertifications, regulatory requirements, and hardware updates. The deadlines are external, which means this work gets priority, whether or not it advances the product.

The shift to NACS is a useful example. What began as a manufacturer-led change moved rapidly toward industry standardization. For every operator running their own platform, that meant connector support, firmware work, session handling and driver communication, much of it necessary simply to keep pace rather than to create a differentiated advantage.

And that work does not arrive in isolation. Metering rules for per-kWh pricing vary by jurisdiction. In Maryland, charging providers spent much of 2025 and early 2026 working out the requirements and fees for compliance with the state’s Weights and Measures Program. Based on AFDC data from August 2026, these changes affected more than 1,700 charging ports across 30+ network operators in a single state.

Publicly funded sites bring their own payment and reporting requirements. Under the federal NEVI program, every state can set additional technical requirements beyond those laid out in 23 CFR Part 680, with rigid compliance timelines. Every new state, manufacturer and program extends the surface the platform has to keep current, while engineering capacity rarely expands at the same rate.

That is why the important number is not simply how many hours this work consumes. It is what those hours prevent the team from doing instead. Mandatory work is easy to see on an engineering board. The product work it delays is much harder to account for.

A crowded engineering backlog does more than slow product development. Over time, it narrows the set of opportunities the business is equipped to pursue. You see it first in the roadmap: a feature slips by a quarter, a commercial capability moves to the next cycle, customer tooling the team knows how to build stays unstarted because mandatory platform work keeps taking priority.

Some of what gets deferred may become standard among the operators you compete with, or a commercial expectation in the segments you want to serve. The deferred roadmap produces no cost line anywhere. It shows up as the capability you lack on the day a commercial opportunity needs it, or as a new segment, partnership or commercial model that stops looking viable once the software work is priced in.

That is what makes this cost so difficult to see. It rarely appears as a failed project or a visible loss. More often, it appears as something the business never got far enough to attempt.

The longer a team runs its own platform, the better it becomes at solving that platform’s problems. What can be harder to see is how other operators are solving the same problems. Without external reference points, a team can overestimate how proprietary some requirements are: billing logic that feels irreplaceable, hardware edge cases that appear uniquely difficult, or workflows that seem too specific for a commercial platform to support.

Some of those requirements really may be unique. Others are part of the ordinary complexity of running a large charging network. Firmware management across manufacturers, multi-tier fleet billing and state-specific payment and reporting requirements are problems many operators have had to solve. A commercial platform used by many operators is exposed to a wider range of hardware, commercial models and operating conditions, which helps separate what genuinely needs to stay proprietary from what the market has already learned to handle.

The hidden cost appears when scarce engineering capacity is spent rebuilding capabilities that feel proprietary internally but are already well understood elsewhere.

An in-house platform can look resilient on paper while depending heavily on a small number of people. Over the years, certain engineers accumulate knowledge that never makes it into the documentation. Why was a billing rule hardcoded? Which firmware update created a particular edge case? Which workaround keeps a specific hardware combination stable?

That becomes a business continuity issue when the people carrying that context leave. Knowledge transfers imperfectly, and the gaps often become visible at exactly the wrong moment: during a major integration, migration, acquisition, or compliance deadline. The relevant question is not whether the team can maintain the platform today, but how much of that ability depends on people the business cannot easily replace.

A commercial platform can spread the cost of developing and maintaining common capabilities across many operators. For an in-house team, each new requirement enters the same internal cycle of evaluation, prioritization, development and maintenance, alongside everything already competing for engineering capacity. The hidden cost is not only what you have to build yourself, but what you have to keep updating yourself year after year.

The costs also reinforce one another. Maintenance consumes capacity that could clear the roadmap, while delayed roadmap work makes it harder for the platform to keep pace with what the business needs next. Over time, the question shifts from which features are missing to which commercial models you can support, and what your engineering team is free to build while the market keeps moving.

Four numbers worth knowing

None of these costs is as easy to measure as a license fee. But that does not make them impossible to assess.

No single number tells you whether owning the platform still makes sense. But the answers to the questions above give you a different view of what your in-house platform is costing the business, not just to maintain, but in capacity, optionality, and time. The In-House CPMS Benchmark works through this assessment in a more structured way. The important question is not whether these costs exist. It is whether they have become material enough to change the ownership equation. The next chapter looks at what some operators chose to do when they reached that point.

Work through the In-House CPMS Benchmark

Why some operators are rethinking what they need to build

So far, this guide has asked whether continuing to build and maintain your EV charging management platform in-house still offers enough advantage to justify the effort. Another useful reference point is what network operators have chosen to keep building themselves, and what they have decided to buy instead.

What follows are recurring patterns drawn from US operators that now use AMPECO. The individual companies have been anonymized, but the experiences come from our direct work with those operators. The value lies in the recurrence: different businesses facing different pressures reached similar conclusions about what they needed to own, what they needed to build themselves, and what they could rely on a specialist platform to provide.

What changed was where they decided control mattered most. The shift was not about handing over the business, but about changing where their own resources and attention were best spent.

For many operators, the trigger is not a strategic review, but a buildup of operational friction: more session failures, rising support demand, manual work piling up, and engineering teams spending more time keeping the network running than improving the business.

Across the US operators we work with, this is a common starting point. Many did not set out to change platforms; they started by trying to solve an operational problem. The broader cost only became apparent when they looked at how much internal capacity the existing model was consuming, and how difficult it would be to keep adding people fast enough to keep pace with the growing complexity.

That is the same pattern Chapter 5 described from the other direction: costs that do not appear in a budget often surface instead in support demand, staffing pressure, delayed roadmap work, and engineering capacity consumed by maintenance.

The second pattern is especially visible among relatively lean teams running increasingly complex networks. As they expand across states, customer types, and funding programs, requirements such as tax treatment, payment compliance, program reporting, and metering accumulate faster than headcount.

The math is similar across these operators. Each new requirement represents a specialist capability that the business must either hire, build, and maintain internally or obtain as part of the platform. When the platform already handles it, the team can direct that capacity elsewhere.

This is most common among operators who are capital-constrained, rather than capability-constrained. They are not short of ideas about what to build. They are short of people to maintain what they have already built. The objective is to allocate less engineering capacity to maintaining the common foundation and more to differentiating the business.

The third pattern applies to operators for whom EV charging sits alongside other responsibilities: property portfolios, multifamily and commercial real estate, hospitality, municipal and university sites, and destination locations. EV charging is a service the business offers, not the business itself. It can be strategically important even when they do not expect to create their advantage by owning the underlying software.

Their requirement is specific and not primarily about features. EV charging has to run without becoming a second business and without introducing a vendor with an interest in the site relationships they already own. That second condition does more work than it appears to: for an operator whose core asset is the relationship with a property owner, tenant, or customer, it matters whether the software provider’s business model could eventually overlap with those same relationships.

Changing course is not the same as retreating

The network operators who reach this decision tend to do so through a structured process rather than by capitulation. One ran a formal evaluation across multiple platforms over nine months. Another assessed several providers and spoke to their existing users before committing. In more than one case, the operator had run an evaluation that had previously yielded a different conclusion and was rerun because something in the business had changed.

Buying was not the fallback when building stopped working. It was the conclusion of a deliberate reassessment by operators that had already invested substantially in how they ran their charging business and could have continued with their existing model.

What none of them gave up

The concern that can stop these decisions is not always cost or capability, but control. But control does not have to mean building and maintaining the platform yourself.

Across these operators, the pattern is consistent, even if the details differ. Some kept their own branded driver apps and data warehouses. Others continued to run proprietary dashboards, middleware or customer portals connected to the commercial platform through APIs. Hardware choice also remained theirs, often across several manufacturers rather than a single ecosystem.

Buying the platform did not mean giving up the parts of the business that their customers see or the things that set them apart. In several cases, engineering capacity shifted toward work visible to drivers or site hosts, while the network operator retained control over its brand, customer experience, data, and commercial strategy.

Different starting points, similar conclusion

One operator was dealing with reliability problems that had become a margin issue. Another was running an increasingly complex network with a lean team. Another wanted charging to work without becoming a software company, while another was preparing for a much larger footprint. Different sizes, segments and pressures led them to a similar conclusion: not that building was wrong, but that for their business at that point in time, continuing to build and maintain the platform internally was no longer the best use of the resources they had.

The more useful question is what your business genuinely needs to build itself to create a market advantage. Your conclusion may be different. The point of the evaluation is being able to explain why. There is one more way to test that decision: ask what happens when the business itself changes shape.

Score the three tests and four numbers in the In-House CPMS Benchmark

M&A optionality: can your software support the business you want to buy?

M&A puts software flexibility to a different kind of test. An acquisition can bring a different mix of chargers, customer types, tariffs, partner integrations, billing models, and operating processes, none of which may fit the assumptions your platform was originally built around.

An in-house platform can run your current network extremely well and still become a constraint when the shape of the business changes. That does not make the platform weak. It means the question is no longer only whether it can support the business you have, but whether it can absorb a different one. That is where software flexibility becomes strategic optionality.

As North American charging markets mature, growth does not have to come only from building new sites. Networks can expand through partnerships, portfolio transfers or acquisitions, while operators can also decide that certain assets are better owned or run by someone else. IONNA taking over operation of approximately 85 Circle K charging sites in the US and Hypercharge adding more than 2,700 ports through its acquisition of Eddie in Canada show what that can look like in practice.

The deal does not end when the transaction closes

Buying a network is one thing. Turning two networks into one operating business is another. Imagine an attractive regional network becomes available. Its chargers need a newer OCPP version than you support. Its drivers register and pay through a different flow. Its B2B contracts follow a commercial model your platform was never designed around.

Could you absorb it without turning the acquisition into a software project?

If the answer is yes, the platform is supporting the deal. If the answer is “yes, but we would need to build…”, the real questions become how much, how long, and what else has to wait.

Software changes the economics of the acquisition

The value of an acquisition usually depends on what happens after closing: consolidating operations, retiring duplicated systems, giving drivers one customer experience, running the acquired network with the same team. A capability gap delays integration, adds engineering work, and postpones the synergies the deal was priced on. In some cases it makes an otherwise attractive network hard to justify. Software does not decide whether a deal is attractive. It decides how much work sits between buying the asset and realizing the value you paid for.

Optionality means being able to act before you know what you will need

Most operators have no idea whether they will acquire another network three years from now. That is exactly what makes M&A such a useful stress test. It asks the platform to take on a business it was never configured for. And the principle reaches beyond M&A: a platform that can absorb an acquired network can also support a new segment, a new business model, or a major partnership. The goal is not to predict what the business will become. It is to preserve the ability to become something different.

Three questions to ask yourself

• If an attractive network became available tomorrow, could your platform support its hardware, customers, and commercial model without a major rebuild?
• How much of the deal’s expected value would depend on software work after closing?
• If your own business were being evaluated, would your technology make integration easier or harder for the buyer?

The strongest software strategy does not tell you which opportunities to pursue. It makes sure the platform is never the reason you cannot. M&A is only one way the business can change shape. The broader question is where the next phase of growth should come from, and where it will actually make the business stronger.

The next growth cycle: scale where it pays, specialize where you win

Most of this guide has been about a decision you already made. This chapter is about one you have not made yet. For much of the last decade, operators could follow broadly the same playbook: raise capital, secure sites, build the network, and position for the demand expected to follow. Growth itself provided the direction.

That playbook is harder to rely on now, as Chapter 1 showed. The question is no longer simply “Where can we grow?” It is “Where does growth make the business stronger?” For some operators, that will mean building greater scale. For others, concentrating on a segment where they can build a defensible position. Some will consolidate, partner or refocus their portfolios. What matters is whether the next move improves the business’s economics and operating position, not just its size.

Utilization is one of the clearest ways that leverage shows up. On a typical fast-charging site, much of the cost base is fixed: hardware, financing, grid connection, and software. Those costs are the same whether the charger is busy or idle, so getting more use from existing infrastructure can strengthen the economics without adding the same level of cost.

Growth strengthens the business only when it improves utilization, efficiency, customer value, or commercial reach. Otherwise, it just adds cost with a bigger number attached.

Operational excellence is what makes either strategy work

Scale and specialization look like different strategies, but they depend on the same foundation: an operating model that gets stronger as the business grows rather than harder to manage. Operational excellence means fewer manual interventions, faster diagnosis and support, better utilization decisions, more effective pricing, stronger retention, and the ability to add complexity without adding cost at the same rate.

The useful test is simple

• If the network doubles, does support demand double with it?
• Does engineering capacity have to rise at the same rate?
• Does every new segment bring another set of manual processes?

Much of that answer is set by the platform underneath the operation. That platform decides how much of the network’s growing complexity becomes automated workflow and how much becomes headcount: whether a pricing change is a configuration or a project, whether a new segment reuses what exists or starts from scratch.

Scale multiplies whatever the platform makes easy, and it multiplies whatever the platform makes manual just as faithfully.

That is why two operators can double the same network and end up with very different cost curves. Operational excellence is not the alternative to growth. It is what makes the next round of growth economically sustainable.

Consolidation is one path, not the strategy

Recent activity in North America shows restructuring takes many forms. A business closes an operation because the economics no longer work. A company sells part of a portfolio to concentrate resources where it sees more value. Two businesses combine assets, or partners restructure who operates an existing network.

Those moves happened for different reasons, but they make the same point. Restructuring is a decision about where capital, assets, and operating attention create the strongest position. It is not a strategy in itself. Buying another network does not fix weak economics, and selling assets does not automatically signal failure.

What has changed is the story investors want to hear. For much of the last decade, the industry could ask them to underwrite what the network might become. Increasingly, the more persuasive story is what the business can already prove: utilization, margins, operating efficiency, and the ability to grow without simply adding cost. That is a harder story to tell. It is also the one the market now rewards.

Use this cycle to fix what the next one would expose

A tighter market can make efficiency feel defensive: protect cash, reduce costs, keep the network running, and wait. But this period also gives operators an opportunity to address the constraints that become much harder to solve once growth accelerates again: retiring underperforming sites, improving support processes, refining pricing, increasing utilization, simplifying operations, strengthening retention, and making sure the technology foundation can support the next segment or business model without another rebuild.

The objective is not to emerge from a difficult cycle intact. It is to emerge with a business that can grow without recreating the same structural problems at a larger scale. The next growth cycle will not erase the lessons of this one. The work now is to make sure the business is ready to grow differently when the opportunity returns.

What will winning CPOs look like in 2030?

Many of the charging networks that will be competing in 2030 are already operating today. Their sites will change, their hardware will evolve, and some will consolidate or move into new segments. But the decisions operators make now will shape what they are capable of by the end of the decade.

The opportunities in this chapter are the reward for getting the basics right, not a route around them. New revenue streams, energy flexibility, and grid participation only matter if the operation underneath them is reliable, efficient, and under control. In many cases, solving today’s problems is what creates tomorrow’s opportunities.

What is changing is the complexity of the job. For most of the industry’s first phase, a charge point operator had a clear task: put chargers in useful places, keep them working, make it easy to start a session, and collect payment. None of that goes away. But the operator of 2030 may also be coordinating charging demand, electricity costs, site capacity, batteries, fleet schedules, customer pricing, utility signals, and commercial partnerships at the same time. The network starts to look less like a collection of chargers and more like an energy and commercial system that has to be coordinated in real time.

No single business model will win. Some operators will scale, some will specialize, and some will become part of larger energy, retail, or infrastructure businesses. What the winners are likely to share is the ability to make more of the assets they already operate, and a platform that lets them take on each of the capabilities below without a rebuild.

The charger becomes only one part of the site

It is still possible to think of a charging site largely in terms of chargers and a grid connection. That becomes less useful as power requirements rise. The IEA’s Current Policies Scenario has total US public charging capacity growing from roughly 13 GW to just under 30 GW between 2025 and 2035.

More power through the same site changes its economics. As Chapter 3 described, a handful of high-power sessions can set the peak that shapes a month’s bill. The value of the site increasingly depends on when and how electricity moves through it, not simply how much it sells.

Software decides when to draw from the grid, when to use stored energy, when charging can be shifted, and how much power each vehicle receives, balancing grid capacity, battery storage, building load, and electricity prices at the same time. By 2030, energy management is likely to move from a specialist capability to a core charging operation.

Pricing becomes part of energy management

Electricity can cost a site different amounts depending on when it is consumed and how much power is drawn at once, while a fleet vehicle may have to leave before the cheapest window arrives. The US Department of Energy describes managed charging systems that adjust power or shift sessions in response to building load, vehicle schedules, utility rates and grid conditions.

That opens the door to pricing that responds to utilization, energy cost and available capacity without exposing that complexity to the driver. IONNA, Walmart EV Charging, bp pulse, and other new operators are already experimenting with promotional pricing and time-of-use rates. As competition for charging dollars increases, balancing affordability and energy costs becomes central to long-term profitability.

AI becomes less visible as it becomes more useful

By 2030, AI may matter less as a feature operators talk about and more as something embedded in how the network runs. Charger status, session history, fault codes, energy use, pricing and arrival patterns can become predictions about failures, demand, schedules, and abnormal behavior.

The distinction will be less between operators that “use AI” and those that do not, and more between operators whose systems produce enough structured, accessible data to automate decisions and those whose data stays trapped across disconnected systems.

AI can identify a problem or an opportunity, but the platform still has to act on it. The value only appears if that insight can trigger a workflow, change a setting, alert the right team, or connect to another system without creating another manual process.

Fleets change what a charging platform has to understand

A driver at a public station needs energy and to get back on the road. Fleet charging is a coordination problem: which vehicles are arriving, how much energy each needs, when they leave, what the depot can support, and what electricity will cost in that window. DOE’s work on fleet charging already treats vehicle operations, facility infrastructure, energy management, and utility signals as one system.

How large that market becomes in the US is unusually uncertain. The IEA’s Current Policies Scenario sees around 40,000 heavy-duty charging points by 2035; its Stated Policies Scenario pushes that past half a million, mostly at depots. An operator does not need to build for half a million fleet chargers today. It does need to know whether fleets could become part of the business without forcing the platform into a different operating model.

The utility becomes part of the operating model

Historically, the utility supplies electricity to the charging site and issues the bill. Increasingly, it can also provide signals that shape how charging is managed: time-of-use prices, grid constraints, demand-response requests, and, in some markets, incentives for flexibility. DOE’s Alternative Fuels Data Center already lists utility programs that let customers opt in to utility-managed EV charging.

Where the driver or fleet has flexibility, some demand can move. At sufficient scale, those decisions become valuable not only to the operator but to the electricity system. DOE’s vehicle-grid integration strategy describes charging responding to site and grid conditions, with flexibility coordinated through utilities, aggregators, or customer-side systems. Operators may end up managing flexible demand on behalf of vehicles, sites, and potentially the grid.

Keep the next options open

Vehicle-to-grid takes that one step further. The potential is real, but the commercial ecosystem is immature: the IEA identified only 22 vehicle models with stated V2G capability in 2026, less than 1.5% of EV models. V2G should not be the basis of a platform decision today. You do not need to know whether V2G becomes a revenue stream. You do need to know whether your systems have already decided the answer for you.

The advantage is not predicting 2030 correctly

No operator knows which of these capabilities will matter most to its business. A highway network may care most about battery storage and energy procurement. A multifamily operator may need utility coordination and load management. A fleet specialist may build around charging schedules and vehicle readiness. The mistake would be trying to predict all of that now and build every conceivable capability in advance. The stronger position is an operating model that can participate when a relevant opportunity arises.

This is where the platform becomes more than the system that runs the network. It becomes part of the operator’s innovation infrastructure. The operator remains the innovator. The platform’s job is to shorten the distance between an idea and the ability to test, integrate, and scale it.

The operators with the most options will not be the ones who predicted the future best. They will be the ones who did not make today’s platform decisions in a way that closes tomorrow’s doors.

How to know when it’s time to reevaluate your platform

An EV charging platform does not have to fail before it deserves another look. More often, it keeps working while the compromises around it accumulate slowly enough to become part of how the business operates. A manual workaround becomes a standard process. A partner integration takes longer than expected, so the next one gets approached more cautiously. A pricing idea moves down the roadmap because it would require too much engineering work.

None of those moments necessarily justifies replacing your EV charging management platform. But taken together, they can tell you the assumptions behind your current software strategy may deserve to be tested again. That is different from deciding to migrate. You can evaluate the platform now, before a constraint becomes a crisis or a strategic opportunity forces the decision on you.

The warning signs look different depending on where you sit

For the people responsible for growth

It may look like lost time. New propositions take longer to launch. Attractive partnerships carry too much technical work. Entering a new market or segment depends on capabilities still sitting in the backlog.

For the people responsible for the economics

It may look like operating leverage that never arrives. The network grows, but engineering and support requirements grow with it, without an equivalent increase in commercial speed or capability.

For the people responsible for the technology

Growing backlogs, integrations that touch more of the architecture than they should, key knowledge concentrated in a handful of engineers, and more capacity consumed by protocols, payments and compliance.

At board or investor level

It appears as strategic constraint. Can the business enter an adjacent segment, integrate an acquisition, or adopt new operating models without first rebuilding the foundation?

These can look like separate problems. Often they are different expressions of the same one. AMPECO’s existing work on CPMS reevaluation describes how this happens gradually: workarounds become “how we do things,” while opportunities the platform cannot easily support are quietly reframed as things the business has decided not to prioritize. The useful question is not simply whether the platform works. It is whether it is still helping the business go where it wants to go.

Go back to the three tests

Those three tests contain most of the questions leadership teams need to ask. There is little value in adding another checklist at this point. The more useful step is to put numbers behind them.

Four numbers make the discussion harder to avoid

Throughout this guide, four pieces of evidence have kept appearing. They are deliberately simple because the point is not to create a technology audit. It is to understand what the software strategy is doing to the business. There is no universal percentage that makes an in-house platform good or bad. The more revealing question is how each number is moving.

Engineering allocation

What share of engineering capacity went to maintaining the core platform last quarter, rather than building capabilities that differentiate the business? How much went to work that moves the business forward, such as pricing changes, customer requests, new segments, and partner integrations, and how much went to keeping pace with the market: protocol versions, payment requirements, metering rules, state program obligations? If the network is scaling and maintenance takes an ever-larger share, the platform may not be delivering the operating leverage you expected.

Commercial speed

How long did the last commercially important idea take to launch? If the time went into research and negotiation, fine. If it went into changing foundational software, that tells a different story.

Roadmap deferral

How many commercially important items have been deferred more than twice? One delayed feature means little. A pattern of losing priority to mandatory platform work is harder to dismiss.

Strategic flexibility

If you acquired another network next quarter, how quickly could you bring it onto your operating model? You may have no acquisition planned. That is precisely why the question works.

Taken together, the numbers tell you more than a feature comparison can. They show whether ownership is translating into speed and strategic freedom, or whether increasing amounts of organizational capacity are being spent preserving it.

Evaluation does not mean migration

Once somebody raises the possibility of evaluating another platform, the conversation jumps ahead to the hardest version of the decision: migration cost, charger cutover, data transfer, customer disruption. Those are legitimate questions if you decide to migrate. They are not prerequisites for finding out whether you should, and the questions operators are most afraid to ask about migration are usually the reason the evaluation never starts.

There are more than two possible conclusions

Get the CPMS Evaluation Criteria Worksheet