The IoT world is growing quickly, but the ground underneath is anything but stable. Platforms close, companies pivot, and once-reliable ecosystems can disappear overnight, turning working hardware into stranded assets. The idea of one “Forever platform” no longer holds. Modern IoT architects now need to build systems that can be separated as easily as they can be connected.
In a landscape where platforms can disappear without warning, the only true stability is that you own your data and can move it anywhere, anytime.
This post covers the essentials that give companies control over their data and devices: high-fidelity exports like JSON and CSV, migration-ready REST APIs, and audit trails that maintain trust when switching vendors.
Drawing on lessons learned from recent platform failures like Gigaset, Osram Lightify, and Sigfox, as well as new rules like the EU Data Act 2025 and NIST IR 8259A, we outline what it takes to design IoT systems that outlast their providers.
The volatility of the connected ecosystem
The IoT market grew with the promise of endless blue ocean opportunities: new devices, new platforms and an $11 trillion horizon. This optimism did not survive contact with the real world. Today’s landscape is more like a Red Lake: crowded, intensely competitive and littered with platforms that have never found sustainable business models.
Security flaws, brittle architectures, and horizontal platforms that delivered little measurable value drove the industry from expansion to consolidation. As we grew, speed and speed of functionality were most important. With consolidation, survival rises to the top.
Anatomy of a shutdown
If a platform collapses, data exports are immediately incomplete or unusable. Devices that rely on the cloud become a dead end. Some recent failures illustrate how this plays out.
Gigaset (2024)
The bankruptcy of Gigaset, a well-known manufacturer of smart home and communication devices, immediately paralyzed its smart home cloud. Without an open API or local fallback, sensors and cameras turned into e-waste. The hardware still worked, but the cloud carried all the intelligence.
Amazon Echo Connect (2024)
Even tech giants are trimming product lines. When Amazon shut down the Echo Connect, the hardware died too. A small gift card couldn’t make up for it “Too big to fail” does not apply to niche IoT products.
Sigfox (2022)
The collapse of Sigfox, a pioneer in low-power wide-area networks (LPWAN), highlighted the danger of tying hardware to a single network operator. A Sigfox radio can only communicate with the Sigfox infrastructure. When the network failed, entire fleets of equipment had to be physically replaced, a painful blow during large deployments.
These errors highlight that without a planned exit path, even stable deployments can fail immediately.
The psychology of supplier loyalty
Vendor lock-in is as much a business strategy as it is a technical flaw. Providers increase the friction of departure, so the cost of migration remains higher than the cost of staying, even if service deteriorates. You see this strategy in three architectural levels.
- Data lock: Data lies behind formats or dashboards and there is no real mass export.
- Logic lock-in: Rules, triggers, and automations run on engines that cannot be migrated.
- Identity lock: Devices rely on the provider’s CA, preventing redeployment elsewhere without physical access.
A “before marriage” The mindset breaks this pattern. Instead of assuming that the platform relationship will last, accept that it may not be the case and design the system so that the separation is clean, controlled and cost-effective. This way you get autonomy in a volatile ecosystem.
The EU Data Law 2025
From September 2025, EU data law will require companies to give users access to the data generated by connected devices. The law defends itself against vendor lock-in. It states that data must be provided in a structured, machine-readable and commonly used format; exactly the kind of export that IoT systems often lack today. The rules apply to both consumer and business products, meaning compliance is no longer optional for global companies following EU standards.
The gold standard for data export
Many business teams ask for CSV files because they open easily in Excel. But for long-term storage, accurate backups, and smooth migrations, architects need JSON. JSON fits how IoT systems actually work. It allows nested items, lists, and clearly defined data types so that all details of an asset, its history, files, and location stay together.
Most modern APIs and databases also support JSON directly, and migration scripts can load and map data with far fewer errors than with CSV. However, the best way to create a true exit strategy is with a full inventory download. A complete package that preserves data, media files and the structure that connects them. This package should contain both CSV and JSON as well as an offline HTML viewer in case of a shutdown.
API access and integration
A static export gives you a one-time snapshot. An API gives you ongoing, real-time access to your data and allows you to migrate gradually rather than rushing everything at once.
REST is the primary way IoT systems communicate today. It replaces older methods like SOAP because it uses less bandwidth and works better with modern web and mobile apps. For architects planning for long-term resiliency, the quality of a provider’s API is more important than anything else it promises.
A reliable API should include:
- Common HTTP methods such as GET, POST, PUT and DELETE as well as well-known response codes such as 200 or 404.
- Ability to retrieve or update one item at a time instead of downloading entire databases.
- Ability to ask for specific data that allows teams to change their system piece by piece.
Immutable audit trails and compliance
When companies move from one IoT platform to another, the data chain is broken. In the end, teams wonder if something was changed during the transfer or if there was a breach with the old provider. Immutable audit trails solve these problems by storing a secure history of all events up to the point of data export.
An audit log is only trustworthy if no one can edit or delete it. Immutability means that every entry is locked the moment it is written, even to administrators or attackers. Industries that follow HIPAA, PCI-DSS, or SOX rely on tamper-proof protocols. In IoT, NIST IR 8259A also recommends retaining secure records of device access for future audits.
Future trends in IoT asset management systems
The “Red Lake” consolidation we are seeing in today’s IoT market will continue well into the 2030s. Further changes are already becoming apparent.
Integrated data kill switches
New laws require providers to include a simple “data eject” button to package all user data into a clean, portable file based on new international standards such as ISO/IEC 21823.
AI-powered migration
Large language models can understand and translate different data structures, making movement easier JSON based exports into new systems without weeks of manual mapping.
Marginal sovereignty
With smarter edge hardware and on-device AI, systems become less dependent on constant cloud access and less vulnerable if a provider shuts down.
Close
Your operational data contains years of knowledge, testing, compliance work and improvements. It is one of your company’s most valuable assets. Don’t lock it into a system that won’t let you leave clean. True IoT maturity means expecting strong export tools and testing them frequently.
It means asking for API access even if you don’t need it right now because you may need it in the future. It also means understanding that the platforms that truly believe in their value will make it easy for you to walk away.
The irony is that these are usually the platforms where customers stay the longest. Because the question is not whether your IoT platform will change at some point, but whether you are ready for it.
