Lutron Spent 25 Years Building Walls. Then It Helped Tear Them Down.

Based on the research ofBerends, Deken and Hummel, "Dynamic platform strategies in an interconnected ecosystem: Navigating standards and complements in the emerging smart home ecosystem," Research Policy, 2026

The Hub That Only Spoke to Itself

In 1997, the US lighting-control incumbent Lutron launched a smart-home hub that connected exactly one kind of product: Lutron's own. It ran on Lutron's own proprietary RadioRA 418 MHz standard, so a Lutron switch could dim a Lutron sensor-controlled light, but nothing outside that one company's catalog could join. A Lutron representative later explained the appeal in the plainest possible terms: "Because everything is developed and produced in-house, you can be pretty sure that everything works as it should." This is what Hans Berends, Fleur Deken, and Jochem Hummel, writing in Research Policy, call a platform mode built on first-party complements connected through a first-party standard, the "closed" quadrant of a four-mode framework they build from a longitudinal analysis of 97 smart-home platforms. It is the mode almost every pioneer chose, because in a nascent ecosystem it was often the only mode available: before 2000, there were simply no third-party standards worth building on, so pioneers like the networking-protocol startup Insteon had to invent their own low-cost, reliable standard from scratch just to get a product to market.

Apple ran a strikingly similar play more than a decade later, just with the roles reversed. When Apple launched HomeKit, it let outside manufacturers build the actual smart-home accessories, but every one of them had to embed an Apple-licensed authentication coprocessor to even talk to an iPhone. The complements came from third parties; the standard, and the right to certify it, stayed entirely with Apple. It took until a quiet twenty-second aside at Apple's 2017 developer conference for that to change, when HomeKit program manager Praveen Chegondi told a half-attentive room, "Starting with iOS 11, we are now introducing an alternate method to authenticate HomeKit accessories. Now accessories can be authenticated by software!" Two very different companies, two very different eras, the same underlying logic: when you are not sure the ecosystem around you can be trusted, you build your own fence first and worry about opening the gate later.

The Trap Hidden Inside Control

The instinct to stay closed is not irrational. Berends, Deken, and Hummel find a genuine value-creation payoff to owning both the complements and the standard: the quality of integration is simply higher, because the hub, the sensors, and the protocol were all designed together rather than negotiated between strangers. There is a value-capture payoff too, and it is the more seductive one. A closed platform sells both the hub and the complements, and because nothing else connects to it, switching costs for the customer are brutal. That is lock-in in its purest form, and it is exactly why Lutron's early strategy looked, and in many ways still looks, like good business.

But the paper's 97-platform dataset shows that payoff eroding as the ecosystem around any single firm fills in. In the sample, 13 of 18 early startups, or 72%, launched before 2007 using only first-party complements on a first-party standard. Among startups that launched after 2007, once standards such as Z-Wave and Zigbee had matured, that share fell to 10 of 36, or 28%. Diversifying entrants, incumbents from adjacent industries entering smart home for the first time, show the identical pattern at a steeper altitude: 13 of 14 early diversifying entrants, 93%, launched fully closed, against 13 of 29 late diversifying entrants, or 45%. The ecosystem did not punish closed strategies directly. It just made every alternative more attractive, because third-party standards meant a platform no longer had to build every complement itself to offer a rich smart home.

Control is the thing that feels safest right up until the moment it starts costing you the complement base you need to stay relevant.

The Standard That Ate the Walls

The real-world ecosystem gave firms exactly the kind of third-party standard the paper describes, and it arrived with the full weight of the industry behind it. In 2019, Amazon, Apple, Comcast, Google, and Samsung's SmartThings joined the Connectivity Standards Alliance, then still called the Zigbee Alliance, to build a shared connectivity standard originally named Project CHIP. On May 11, 2021, the Alliance unveiled the result under a new name: Matter. The specification itself, plus a working certification program and eight authorized test labs, shipped on October 4, 2022, backed by what the Alliance described as an international community of more than 550 technology companies. Weeks later, Apple folded Matter support directly into iOS 16.1, with its own release notes stating plainly that the update added "Matter, the new smart home connectivity standard, ... enabling a wide variety of smart home accessories to work together across ecosystems." The same company that had spent years requiring its own authentication chip now pointed developers toward a standard it did not own and could not fully control.

The detail that makes this more than a coincidence of timing is who signed up to build Matter in the first place. Listed among the Connectivity Standards Alliance's own board of directors in that 2021 announcement, alongside Amazon, Apple, Google, and Samsung's SmartThings, is Lutron Electronics: the same firm that spent 1997 through the 2000s proving a closed, first-party standard could win on integration quality alone. The pioneer of the closed strategy is now a governing member of the body that built the open one. That is not a contradiction in Lutron's strategy; it is close to a textbook illustration of what the paper calls combining modes rather than abandoning them, keeping a premium, tightly integrated first-party line while also meeting the ecosystem where it is now headed.

Why Nobody Fully Lets Go

What is easy to miss in the "closed is losing" headline is that almost nobody in the paper's sample went fully open either. Firms increasingly combine platform modes rather than replace one with another, and the combinations are not random: they compartmentalize, keeping the value-creation and value-capture benefits of each mode for a different slice of the product line. Control4, a US startup founded in 2004 by a team that included former Google CEO Eric Schmidt, spent its first eight years refusing to connect anything that was not Control4's own hardware on Control4's own standard. In 2012, with far more third-party devices on the market, its CEO Martin Plaehn reversed course: "We love it that there are all these new devices. Besides the extra attention these newcomers bring to the industry, they also bring innovation and hardware that can only enhance the Control4 ecosystem." By 2014, Control4 had added support for the "Works with Nest" standard specifically so customers could use the popular Nest thermostat, while keeping its own standard for its core product line.

Loxone, an Austrian startup, tried the opposite order and got the same answer. It launched in 2010 intending to sell a fully closed, high-end integrated system, then discovered within a year that it needed to connect to other companies' products to convince people to buy in at all. "We like to say we are the most open system there is, which is why a lot of people buy us," its CEO said, "At the same time, this also benefits the sales of our own high-quality, specialized products." Loxone's revenue doubled every year after that shift, reaching €45 million by 2015. Fibaro, a Polish platform also launched in 2010, never bothered with the closed phase at all, building on open standards from day one specifically so it would never be "reduced" to selling only its own products. Three different starting points, three different timelines, and the same conclusion: the complements you make yourself and the standard you control still matter, but only as one piece of a portfolio, not the whole strategy.

The Rule for Anyone Who Owns a Standard

A proprietary standard looks, on paper, like the most defensible asset a platform can own: nobody else can take it from you, and nobody else's product works without your permission. The paper's finding, across almost three decades and 97 platforms, is that this defensibility is exactly what makes it a trap. The more tightly you control the standard, the more completely you also cap the complements, the user base, and the partners who are willing to build around you, and in a market where a third-party standard is available, that ceiling gets more expensive every year you refuse to raise it. Lutron did not survive by staying closed; it survived by recognizing, early enough, that sitting on the board that builds the open alternative was safer than being the last closed hub standing. Any platform still weighing whether to protect its own standard or open up to someone else's should ask the question the data already answered for Lutron: is your proprietary standard still buying you integration quality worth losing the ecosystem over, or is it just a wall you built before you knew how big the neighborhood was going to get.