MWC is a brilliant show, of course. Not because of the big talk and shiny displays, although they are fun, too. But because you can get lost on the main streets and back alleys of this pop-up tech metropolis for a week at the end of February; because somewhere in Barcelona, between the flashes of late winter sun, you can find whatever you want. Telco? Follow the lights, sir. Handsets, modules, chips? This way, my man. Compute wizardry? Yeah, scattered all about. IoT, you say? Out back. Enterprise? Everywhere you look, and nowhere at all. AI? Are you taking the mick?
It is funny to talk to the old (low-power section of the) IoT crowd about MWC, especially the stalwart non-cellular elements that made IoT a long-range connectivity discipline in the first place. It used to be that you’d come across them in halls seven or eight (out back) after a couple of days’ getting fried in the hot air and UV buzz of the main stands; they’d be leaning against painted MDF booths, looking across empty space, a little forlorn. But what they said was most interesting – about private networks, ecosystem building, industrial revolution.
Which, of course, has become the mantra in the main halls as well over the last few years – as mobile operators have struggled to monetise 5G services and make money back on 5G investments, and as spectrum regulation in global markets has opened local airwaves to private cellular, and created a noisy battlefield for their awkward offensive in the enterprise space. It is a complicated scene, which the non-cellular IoT crowd has been navigating for a decade – and appears to have finally crossed, perhaps, in mob-handed fashion.
Indeed, these high-power IoT execs might do well to swing by the low-power IoT mob in hall eight, and prick-up their ears and fill-up the space. Because IoT has been a chastening journey, so far; great fortunes have been lost, and hardly made. But those that have survived are humbler, somehow, and also united – as if they owe their existence to a tight-knit company of tech soldiers. They are resolved, finally, that they know how to make a decent and impactful business out of IoT, if not an easy one. That seems to be the message, at least.
It is how Wienke Giezeman, chief executive at The Things Industries (TTI), the Amsterdam-based LoRaWAN collective selling core networking and tooling for low-power IoT systems, puts it. Kind-of. He has phoned for a chat, just because he likes to chat – and because he’s good at it, and he thinks the IoT space has rationalised (“not consolidated”) and the big MWC shindig is a diary-marker for the whole market. So what’s the lesson? “That there is no fairytale in IoT – about one-stop shops, and easy wins,” he responds. It goes for high-power IoT, too, surely?
“I don’t know. But IoT is hard, and the companies that are successful are the ones that know how to cherry-pick. Successful IoT solutions comprise a stack, and the simpler the stack, the better,” he says. But the same narrative is being peddled in the private 5G market; that 5G has gone from being the king of new digital tech to being a servant (among many) of new digital enterprise, and talk about cure-alls and ‘primes’ has been swapped for niche pursuits and team-plays. There must be lessons for high-power IoT in low-power IoT, which can cut through all the MWC BS?
That would make a great keynote, suggests RCR. “I think Afzal [Mangal] from Deutsche Telekom has already given that talk,” responds Giezeman, referring to the founder of the German operator’s IoT developer startup unit, IoT Creators, and its former head of IoT marketing and innovation, and a familiar on the LoRaWAN circuit besides. He goes on: “He has talked about the end of telcos, and why they’ve struggled with new business. But the story is a bit late; they have embraced their roles as bit pipes.” It’s not quite the same, but Giezeman is not overly-interested in the private 5G market, anyway.
He prefers to talk about what has gone wrong, and also right, in everyday IoT. “Contrary to the standard analysis, there is no consolidation – really it is that the initiatives that don’t work just stop. And smaller companies quit, and merge, and get swallowed-up as a result.” But there are lessons here. The IoT market remains fragmented and complex, he says; there aren’t a handful of modules for a handful of phones, which all do the same thing. TTI supports 1,500 different devices off-the-shelf, he says – to plug into all kinds of enterprise efficiency projects.
He talks about putting total focus on component pieces and constant innovation, so the pieces are “cherry-picked” by the developer for the total solution; he talks about “sceptical opportunism” as the only way to navigate between the hype and reality. “If you embrace this mindset of sceptical opportunism, then the future is bright. Because you have to be sceptical – about what people say versus what they deliver. We have seen that with Sigfox and NB-IoT – and with LoRaWAN, as well. The whole market has over-promised,” he says.
“But there is this undercurrent of demand, disconnected from the hype, which just goes its own way and selects from the best IoT firms. The enterprises and vendors that are successful are just great cherry-pickers.” Demand is driving IoT revenues up by 25 percent per year, he reckons. “Which is not bad, right? It is still more than the telco market.” Connection volumes, meanwhile, are spiralling upwards by 100 percent. But margins are slim, clearly – and slimming further as the volumes jump. The economics do not stack up for venture firms, bailing fast on IoT, he suggests.
“Bottom line, revenue growth is about 25 percent in IoT, which is good, but not good enough for the venture capital (VC) market; and profits are marginal. So you have to outgrow the market, and very few companies have managed that.” So here is the thing, then – in early 2024, ahead of some kind of nominal MWC reckoning; venture money and marketing hype, chicken and egg, have created an IoT market that never was – just because IoT, which mirrors every enterprise it serves as a sensor-based digital twin, cannot be shoehorned into an easy sale or a single market.
And so VC funds have stopped, more or less, contends Giezeman, and IoT companies have vanished, more or less. “VC strategy is to invest in a lot of companies in the hope that one will succeed, and to invest hard in order to give them a competitive advantage; to have a wide spread, and to go big. But the true economics haven’t worked in IoT. In a lot of VC investments, it is fine to fix the unit metrics later. But if you don’t grow, or you don’t fix them in the mid-term, then you’re going to fail. Because you are going to run out of money.”
Sigfox, which raised at least $311 million in five venture rounds in the period to 2016, is the most obvious case study; it went bust in 2022, before being rescued by Taiwan-based Unabiz (which continues to raise venture funding, it might be noted). There are plenty of other VC-backed IoT startups, although they would bridle at being presented as case studies in Giezeman’s cautionary tale. But perhaps US-based Senet, which has raised about $45 million, might be put forth, just because it sold to Sweden-based LoRaWAN group Netmore last month. Giezeman doesn’t want to say.
But he says: “Lots of IoT companies raised lots of money. Lots of LoRaWAN companies raised lots of money. And the [cellular] operators have spent so much, as well. Half a billion has just been burnt, and there is nothing left. It’s gone. All the money has gone. One thing is clear from the last five years: that raising money has not been the answer… Because it drives this irrational need for diversification – to sell more products to address more markets, to raise more funds. But you can’t be good at everything in IoT.”
More lessons. He goes on: “It is really hard to combine a horizontal technology competence with a vertical technology competence. That is just incredibly hard. Like I say, the IoT companies that have been successful have focused on one business problem and one vertical – with an IoT stack cherry-picked from the IoT toolbox. Or else they’ve supplied the best tool in that box. Everything in between – platforms, middleware, other bits; one-stop shops – is not going to fly. And if the demand lowers a bit, and the liquidity lowers, then you’re going to be in trouble fast.”
There is more – which suggests the point that the old IoT market has reached some kind of maturity, and might be in a reasonable position to preach to the new IoT market. “Everyone is starting to focus on what they do well, instead of resorting to a Powerpoint contract that says they can do everything – just to expand their addressable market, and make venture capitalists interested in them.” The personal proof for all this – which Giezeman wants to share, clearly – is that TTI has dodged the VC bullet, and is outrunning the market with growth of 60 percent.
“We are winning from all of these other companies that have raised money, and come upon hard times. Last year we came on par [at 25 percent growth], and this year we have moved beyond [at 60 percent] – driven by a wide range of use cases with high profit margins. We are growing because our customers are growing. And we are profitable; we have an 80 percent margin, in line with the SaaS model – which goes back into R&D to deliver a better product every month on a ‘hot’ upgrade, in the background without any downtime.”
We shuffle through a couple of metaphors to make sense of the VC bust, and what it really takes to succeed in the IoT space. So is it like what they say about the British surf; that if you can swim out in the choppy waves of a Cornish winter, then you can hang-ten anywhere – but that nothing prepares you for the return trip? “Yes. Because venture money forces you on to big waves, before you know how to do it. And when you’re committed to scale, you don’t have a lot of scenarios left. Most of our time goes on worrying about how to make a better product every week.”
Right, so is the VC effect on an IoT startup like a big record label snapping up the hottest act on the Indie scene, and ruining its sound to make it work for the pop charts? “Kind of; the same dynamics. You’re contracting a lot of artists, and only a few will make it. But this market, adopting sensors and telemetry, is a real market; it just doesn’t grow that fast. Twenty-five percent is a lot, still – compared to most markets. But it is not enough for VCs, and if you have that mismatch – a label that doesn’t match with the band, and vice versa – then you just get friction.”
He adds: “VCs are star makers, and they sit on the board, and, at some point, only worry about getting their money back. The most hurtful thing for IoT companies is to put on a corporate jacket too early. Whereas we have been agile and matured, and grown our corporate profile – not by raising money, but by making the best product in its category. Which is not to say the time won’t come, either. I am not excluding VC. I am just saying that for the last five years, it has not been a good match. We can do it any time. It is a strategic option, always. But that is hindsight, right?”
RCR still thinks Giezeman, or others, should take that headline slot at the MWC keynotes. In fact, later in the conversation – reflecting on the hyperactive 3GPP release schedule, as the hype about 5G-based RedCap starts to mess with the 4G-based NB-IoT/LTE-M sales agenda, just as the latter starts to settle – he outlines a good one. “What enterprise IoT hardware makers want is stability. This is not a market that upgrades every year. We put devices in the field for 20 years; the minimum is four. Because the business case is very hard otherwise.
“And this is a fragmented market. It doesn’t just comprise four module makers and four handset makers. With IoT, there are thousands of different implementations. So these people don’t want an innovation roadmap, actually. It sounds counter intuitive, but, as a device maker, that 3GPP roadmap scares the hell out of me. Because as a device maker, I want to reduce the permutations and combinations. I need fewer SKUs; because otherwise my overheads explode. So, yeah, that roadmap terrifies IoT device makers; that would be a great presentation.”
Giezeman also – finally, sort of – suggests a lesson for how to roll the dice on private 5G. “You should explore IoT in its traditional sense, first. Go with the lowest barrier to entry, which is LoRaWAN – and then, once you’ve gone on the learning curve, take a look at 5G as a stepping stone or a hybrid solution. But while there is big potential, the 5G ecosystem is very immature, and just too much of a leap. We have nearly two million devices on our platform, and it works; the business case is solid. So if you have money to burn on enterprise IoT, then start with LoRaWAN.”
