When the best product is not enough: the Cambium Networks case

How a company able to build excellent gear can lose to competitors that are larger, more visible and sometimes technically more limited. The Cambium Networks crisis tells us far more than Cambium: it tells us what the technology market has become.

There is something deeply interesting, and at the same time worrying, about what is happening to Cambium Networks.

Because it would be much simpler to tell the story of a company that made mediocre products, lost customers, piled up debt and finally fell into crisis. That would be an almost natural dynamic: you make worse products than your competitors, and the market punishes you.

The problem is that Cambium Networks is not that.

Anyone who has actually worked with its radios, with the PMP and PTP platforms, with ePMP, with many of the cnPilot access points and with the whole cnMaestro management ecosystem knows that Cambium has built, and keeps building, technically excellent gear.

Not perfect. No vendor is.

But often designed with that telecom-company mindset in which stability, radio behavior, spectrum management, reliability and predictability come before the visual effect.

And yet today Cambium is in a crisis that is no longer merely commercial.

It is an existential crisis.

The 2023 breaking point

To understand how we got here we have to go back to the period right after the global semiconductor crisis.

During the years of broken supply chains, distributors and operators had learned a very simple rule: if a product was available you had to order it, because nobody knew when the next one would arrive.

Lead times had stretched, components were scarce, and the channel's natural behavior had become stockpiling inventory.

When, between 2022 and 2023, the production situation began to normalize, that huge safety buffer suddenly turned into a problem.

Cambium explained it openly in its financial filings: distributors had increased their purchases during the supply crisis and later found themselves with too much stock, especially in the Enterprise division. In 2023 they therefore drastically cut new orders and started to run down existing inventory. Cambium even had to grant special incentives and discounts to speed up the reduction of channel stock. In the fourth quarter of 2023 alone these initiatives reduced reported revenue by about 11 million dollars.

At the same time, one of the advantages Cambium had enjoyed during the shortage disappeared: being able to deliver when other vendors were in serious difficulty.

It is Cambium itself that writes that, with components available again, its previous advantage in supply and delivery capacity had shrunk. On top of that came very aggressive competitor pricing and weaker demand.

From there, the decline.

After the subsequent restatement, Cambium reported revenue of about 226.8 million dollars in 2023, falling to 177.3 million in 2024 and further to 159.6 million in 2025. In two years that means losing about 30% of revenue, starting from a situation that was already badly deteriorated compared with previous years.

2025 closed with a net loss of about 38.5 million dollars, after a loss of more than 74 million in 2024. At the end of 2025 available liquidity had fallen to about 11.3 million dollars.

Meanwhile the problem had become financial.

In June 2025 Cambium had about 21.5 million dollars of term loan and another 45 million drawn on the revolving credit facility. It had breached its financial covenants and had stopped making the required principal and interest payments on the term loan and the interest on the revolving facility. The lender therefore had the right to demand immediate repayment of more than 66 million dollars.

From that moment we are no longer talking only about a vendor with a bad quarter.

We are talking about a technologically sound vendor that is financially trapped.

In March 2026 Nasdaq suspended trading in Cambium shares. In August the Listing Council confirmed the delisting decision.

Then September arrived.

On 11 September 2026 Cambium cut 260 jobs worldwide, equal to 53.6% of its entire global workforce, with immediate effect. A few days later Cambium Networks Ltd, the group's British subsidiary, formally entered administration in the United Kingdom. The administrators are trying to sell assets and some or all of the business lines involved in the procedure, while liquidation through the relevant local procedures is expected for the group's remaining CNL entities.

This is not a normal restructuring.

It is the dismantling of an industrial structure.

But if the products were good, why did the market not save them?

This is probably the most interesting question.

Because it contains a belief we keep holding in the technology world: that in the end the best product must win.

It does not.

In fact, it probably never did.

A technically excellent product can lose to one that is more limited, more closed, sometimes even more expensive, if the second has a stronger commercial ecosystem, more visibility, a wider sales network and a better brand perception.

In networking this shift has become blatantly clear.

For decades a technician evaluated an access point by looking first at the radio, the behavior under load, reliability, roaming, the ability to handle interference, multicast, VLANs, authentication, QoS and difficult situations.

Today, more and more often, the first assessment happens by looking at a dashboard.

The interface has to be beautiful.

The app has to look modern.

The chart has to be animated.

The packaging has to convey a premium feeling.

Onboarding has to take three clicks.

And above all the product has to be tellable.

This is not necessarily a bad thing. Simplicity is a quality, and a good interface is an important part of a good product.

The problem arises when perceived quality replaces technical quality.

It is a cultural shift that the consumer technology market has hugely accelerated.

Apple, for better or worse, taught the whole industry that a technology product is judged not only for what it does, but for how it is presented, touched, configured and perceived.

Apple, however, generally paired this construction of perception with enormous investment in product engineering, hardware, software, integration and supply chain control.

Many other vendors mostly learned the easier part of the lesson.

The aesthetics.

The dashboard.

The minimalist design.

The apparently simple ecosystem.

The marketing message that all the complexity has magically disappeared.

But a network is not a smartphone.

The physics of radio waves does not become simpler because the interface has rounded corners.

Cambium ended up in the worst possible spot

Meanwhile Cambium found itself squeezed between two worlds.

At the more affordable end of the market, Ubiquiti has built an ecosystem of gigantic size. You do not need to argue that every Ubiquiti product is technically better than an equivalent Cambium one to understand why the model succeeded.

It is enough to look at the numbers.

In the fiscal year ended June 2025 Ubiquiti reported 2.57 billion dollars of total revenue, of which about 2.25 billion from the Enterprise Technology platform. In the same year the total revenue of all of Cambium Networks was about 160 million dollars. The categories and fiscal years are not directly comparable, but the difference in scale is striking.

That scale changes everything.

It lets you buy components on better terms, fund more product families, absorb design mistakes, sustain periods of lower margins, develop software, produce enormous amounts of documentation and content, build communities, distribute gear almost everywhere and, above all, become a familiar choice.

Ubiquiti has also progressively built a much more direct relationship with the customer: in FY2025 about 44% of revenue came directly from its own webstores, while the remaining 56% came through distribution.

Cambium, on the other hand, kept depending heavily on a traditional structure of distributors, resellers and integrators.

When that chain gets too full of product, the vendor does not immediately see the drop in end demand. It keeps shipping, thinking the market is absorbing the gear.

Then suddenly the distributor stops ordering.

And that is exactly what happened.

On the other side of the market Cambium had to fight against the likes of Cisco, HPE Aruba, Juniper, Fortinet and other large enterprise vendors.

Here the problem became the opposite.

The customer was not just buying an access point.

They were buying global contracts, firewalls, switching, NAC, SD-WAN, security, enterprise support, certifications, consulting and commercial relationships built over decades.

So Cambium faced competitors able to win on price and simplicity and, at the same time, competitors able to win on size and commercial weight.

It is an extremely difficult position for a relatively small vendor.

The Meru Networks precedent

Anyone who has worked in enterprise Wi-Fi long enough has already seen something similar.

Meru Networks.

Meru was one of those vendors that made engineers fall in love and often drove salespeople crazy.

It had developed a Wi-Fi architecture profoundly different from the traditional one, built around the concept of Virtual Cell and single-channel architecture. It was particularly interesting in high-density environments and in applications where mobility and continuity were fundamentally important.

It was not simply "another access point".

It was a different vision of Wi-Fi.

In 2015 Fortinet acquired Meru for an announced equity value of about 44 million dollars. The stated goal was to integrate that technology into Fortinet's Secure Wireless strategy.

Formally, Meru did not vanish the next day.

Its technology kept living inside the FortiWLC platform and part of its technical heritage remained available for years.

But the transformation began immediately.

In the quarter after the acquisition Fortinet explicitly stated that it had started cost-reduction and restructuring activities tied to Meru, including workforce reductions, real-estate consolidation, asset write-downs and contract terminations.

The former Meru platform thus became one of the components of Fortinet's huge portfolio.

Its single-channel architecture kept existing inside FortiWLC, but it no longer represented the identity of a company built entirely around that vision. Even years later, Fortinet's own community documentation identified FortiWLC as the platform derived from the Meru acquisition and explicitly distinguished its technology from the other Fortinet wireless management modes.

Progressively, many original Meru devices became legacy. Fortinet confirmed that several Meru systems are no longer supported and that the related legacy licenses are no longer available.

This is how the industry normally works.

An extraordinary technology can be bought not because someone necessarily wants to preserve its identity, but because patents, developers, customers and intellectual property have value within a larger strategy.

And Cambium's story could take a similar direction.

We do not yet know who will buy what, nor whether the different Cambium lines will stay together.

But we already know that the administrators are looking for buyers for assets and business lines.

PMP might interest one party.

ePMP another.

Enterprise Wi-Fi and switching might end up elsewhere.

cnMaestro might have yet another fate.

And at that point Cambium Networks, understood as that particular combination of people, design philosophy, radios, software and technical culture, might simply no longer exist in the form in which we knew it.

cnMaestro is perhaps the most bitter symbol of this story

There is then a particularly significant aspect.

cnMaestro is probably one of the best demonstrations that Cambium had understood perfectly where the market was going.

A single platform from which to control access points, switches, fixed wireless links and different families of devices spread around the world.

Inventory.

Configuration.

Firmware.

Statistics.

Alarms.

Monitoring.

Provisioning.

A Cambium network could be geographically scattered over hundreds or thousands of kilometers and still appear to the administrator as a single system.

It was exactly the right direction.

And now that very centralization risks turning into the biggest problem.

Cambium has announced that cnMaestro Cloud may no longer support Enterprise devices after 1 October 2026 and is advising customers to migrate Wi-Fi access points and switches to cnMaestro On-Premises 6.0. It is not correct to claim today that all of cnMaestro Cloud will certainly be shut down on the first of October: for Fixed Wireless, for example, Cambium still has to give definitive guidance. But the very fact that a vendor is inviting its worldwide base, within a few days, to move Enterprise management from its cloud to standalone servers makes the seriousness of the situation clear.

It is hard to imagine a more concrete demonstration of how quickly the meaning of the word "cloud" can change.

Yesterday it meant not having to install servers.

Not having to update controllers.

Not having to manage backups.

Being able to reach thousands of devices from anywhere in the world through infrastructure maintained directly by the vendor.

Today some of those same customers have to quickly install virtual machines, export configurations, change the devices' reference controller, rebuild missing functions and check that gear located maybe in remote sites can still be reached.

Cambium even stated that, should the Cloud Anchor be disconnected, the current cnMaestro On-Premises would keep working but would lose the ability to add and remove devices. The company therefore plans, before any shutdown of the Anchor, to make available for Wi-Fi and switching an On-Premises image free of the dependency on the Cloud Anchor and on licenses.

It is almost the complete reversal of the paradigm with which the industry sold us the cloud over the last fifteen years.

And this is where the Cambium affair stops being just the story of Cambium Networks.

It becomes the story of a sector that has slowly handed over not only its data, but often the very operability of the infrastructure, to the vendor's servers.

It is a huge topic, one that deserves a separate discussion.

Because when we talk about a network with ten access points the problem is annoying.

When we talk about an operator, a university, a hotel chain, a public body or an MSP with thousands of devices spread across the territory, the disappearance or the change in the vendor's cloud policy becomes an infrastructural issue.

And it does not concern only Cambium.

The technical defeat that is not technical

So the most bitter lesson of the Cambium affair is a different one.

We can keep telling ourselves that the market always selects the best product.

But that is not what happens.

The market selects a combination of technology, price, distribution, marketing, availability, ecosystem, brand perception, commercial relationships and financial strength.

Technical quality is only one of the variables.

Sometimes not even the most important one.

A device can have a better radio, more serious engineering and a longer operational life, and still lose to a product that arrives tomorrow morning from the distributor, costs less, has a more elegant dashboard and is already used by every installer in the area.

Or it can lose to a more expensive and not necessarily technically superior product, because that product carries the name of a large vendor, a global enterprise contract and the perceived peace of mind of buying from a company nobody imagines could disappear.

Cambium ended up right in the middle.

Too professional to be the easy-to-sell tech toy.

Too small to be the enterprise giant impossible to ignore.

And perhaps too engineering-driven in a market that has learned to sell the appearance of simplicity better and better.

This does not mean that marketing and design are useless.

On the contrary.

It means that anyone who builds technically excellent products and thinks their quality is enough to sell them is making a mistake.

History is full of magnificent technologies that lost.

Meru is one example.

Cambium risks becoming another.

And perhaps this is the most frustrating aspect of the whole affair: watching a vendor able to build gear that keeps working for years, designed by people who clearly really understand networks and radio frequency, being progressively dismantled not because it suddenly forgot how to build a good product, but because being good at building a product and managing to win a market have become two completely different activities.

The first problem is solved by engineers.

The second is solved by scale, capital, distribution, marketing and commercial power.

And today, unfortunately, it is often the latter that decide which technologies we will keep using tomorrow.