The story of IoT has never been a simple one. Few technologies have inspired such sweeping visions of transformation, only to stutter and splutter, and fizzle out. For years, the industry has been caught in a cycle of overstatement and retrenchment. But a period of recalibration in 2025 sees IoT enter a different phase — one defined less by promise, and more by experience and results.
For the past decade, IoT has existed in a strange land, somewhere between ultimate promise and total disappointment. In the early days, expectation ran way ahead – so far ahead! – of execution; and then, there were long years of write-offs and bankruptcies, and accelerating market consolidation. It left many wondering whether the original vision – about instrumenting enterprises, saving the planet, making some money along the way – would ever materialize. But in 2025, something finally shifted.
Not because of a single breakthrough or announcement, but because the industry, without fanfare, effectively corrected itself. Little by little and piece by piece, the things that were broken in IoT were chucked out; the things that worked were retained. And what is left? An IoT market whose precise execution – about the negotiation, commission, deployment, management of sensor solutions for business problems – outruns all of its big talk and grand promise. It is an industry that is now completely focused – we would like to argue.
At The Things Industries, we operate thousands of private low-power IoT networks across virtually every imaginable vertical. Our customers use The Things Stack Cloud to connect edge gateways and ultra-low-power sensors, collecting data that flows reliably from physical environments into operational systems. The use cases are often profoundly unglamorous, but they are also deeply valuable – condition monitoring in commercial buildings, temperature tracking for food compliance, cattle tracking on dairy farms, smart water metering for utilities.
To name just a few use cases and applications. The point is that these are no longer pilot systems; they represent essential componentry and critical infrastructure. When analysts ask about growth – year-over-year, as a common gauge of market progress – they underestimate how difficult that metric is to apply in an industry with sales cycles that routinely exceed 12 months. Yet despite that, the IoT industry has (without fanfare) demonstrated something remarkable: annual growth of 20 percent, year after year, for five years straight.
It might not be explosive growth, but it is consistent and persistent growth. And importantly, it demonstrates the new business resiliency of the sector.
Something else: this mismatch between hype and reality, which the IoT market has struggled to shake for so long, is behind us. The buzz about the potential adoption and impact of IoT has never aligned with the reality; subsequent ‘troughs of disappointment’, which invariably follow on the heels of the hype, have underestimated the progress in the market. It has been rain or shine, boom or bust in IoT for 20 years; a case of inflated hopes or deflated expectations, followed by incremental advances that have only become clear in hindsight.
In 2025, that progress has been made clear; the market has righted itself, in line with its rate of delivery. There is less spectacle, more substance; the market is finally moving in step with real-world industrial change-cycles, and not just with self-serving marketing strategies or hyper-active private equity agendas. The promise is still the same, but the delivery is completely different. The market is way healthier than it appears, or than commonly characterised. And the funny thing is that enterprises never cared for, or heard much of, the IoT industry chatter in the first place.
From an end-customer perspective, the reality is that thousands of vertical-specific use cases are scaling simultaneously. Enterprises are integrating IoT into their core operations – not as experiments, but as critical parts of how their businesses function. This is the market many envisioned 10 years ago. It is finally taking shape in a way that is ready to scale. And as 2025 turns into 2026, what makes this moment so interesting is that it is unfolding almost entirely outside of the AI spotlight.
AI dominates everything – market attention, investment capital, column inches. And it does this with no certainty about proper timelines, valuations, or returns. In fact, there is only one thing that is not uncertain about AI: its extraordinary hunger for data. Which is where IoT comes in, of course. As my friend Allen Proithis (at Capstone Partners) put it at The Things Conference in September: at heart, “IoT is a data facilitation business and AI is the most hungriest consumer of data the world has seen. And IoT is the provider of that data”.
Industrial IoT systems generate exactly the kind of data that AI needs: structured, continuous, contextual, and proprietary. Data that belongs to enterprises is grounded in physical reality, and cannot be easily replicated. IoT may still be hard, but the pace at which the market is now forming around real end-user value is striking. And despite being deeply incentivized not to create the next hype cycle, I am confident that in 2026 the broader market will begin to recognize what is already visible in the data today: that the inflection point has passed.
At The Things Industries, we see this shift clearly in the data. Today, more than four million devices are connected through our platform worldwide, routing roughly four billion data points every month. Over 250,000 developers used our platform to explore and build low-power IoT solutions, while thousands of vertical solution providers deploy and operate networks across industries ranging from utilities and agriculture to logistics, buildings and manufacturing. This scale provides a practical view into where IoT adoption is real, repeatable, and ready to grow.
The most important outcome of this shift is not growth alone. It is margin. And not just profit margin, but margin for error, allowing systems to scale reliably; and also margin in timelines, enabling better planning and deployment, and margin for research and development, so companies can move from project-based survival to platform-based innovation. And importantly, margin for collaboration – where partnerships replace zero-sum competition with customers.
That margin is what turns technology into infrastructure. And infrastructure, once in place, has a way of quietly reshaping industries long after the hype has moved on. Maybe 2026 is going to be the year we really stop talking about IoT, and start talking about data pipes – or Data Facilitation Systems. Allen? Are we ready for a new acronym – industrial private DFS, formerly IoT, formerly M2M? Or maybe we should just carry on solving enterprise problems, driving operational improvements, and growing this market as a critical enterprise infrastructure?!
