Although the smart home has become the latest battleground for juggernauts such as Google, Amazon, and Apple, it is still early days for a young market that is also being pulled in all directions. Given the nascent market landscape—and despite the mind-boggling fragmentation of industry players—the road map in the future for smart homes is becoming clear, with solid growth to continue for the market and a distinct emphasis on machine learning as the way forward.
Global revenue for the smart home market is forecast to reach approximately $18.0 billion in 2018, nearly double the $9.8 billion reached two years earlier in 2016.
However, for this growth to be realized, industry players including manufacturers, installers, and service providers will need to make incremental modifications to strategies surrounding pricing models, ease of use, privacy and security, and machine learning.
Already, pricing models have begun to change at the device level with a wide spectrum of options, especially around smart lighting, now available. Philips Hue had been the de facto smart light bulb provider in the past, but it now is part of an increasingly crowded market segment populated with companies including Lifx, Cree, C by GE and Ikea. And notwithstanding pressure from dozens of other players, it is unlikely that premium brands will change pricing models because companies such as Philips have an entire ecosystem of lighting devices to date, even if it had no recurring-monthly-revenue (RMR) model to support development—hence, the higher device price point.
For professional services, providers have made concessions permitting do-it-yourself (DIY) installation alongside professional monitoring—a significant change from past business models, especially from mainstays like ADT, which now has Canopy (never mind that it has yet to materialize, even though the first announcement was made in 2016).
Meanwhile, other professional service providers offer options for financing hardware. This allows a consumer to eventually own the hardware, effectively lowering the monthly bill after the equipment is paid for—which is another big change for professional security monitoring.
Overall, pricing plays a significant role in the path to mass adoption because most smart home devices are priced out of the range that many consider would qualify as an impulse buy. IHS Markit expects that until pricing reaches this impulse-buy threshold, most large-quantity purchases will be reserved for home construction and home renovation projects; or through professional service providers like Utah-based Vivint, Moni from Dallas, or US-wide provider Comcast, which reduces the market opportunity for device manufacturers that could be obtained through the retail channel.
To be sure, the smart home today is more about point solutions that solve a specific problem—not about installing dozens of devices simultaneously. To get to the mass market, smart home solutions need to be simple to set up, and there must be more clarity about which devices are interoperable. This is where professional service providers excel in terms of delivering a positive smart home experience, while the retail channel has enjoyed less success in comparison. For the smart home market to reach mass adoption, viable DIY options must be present for those not willing to shell out monthly fees or resort to using professional services when all that is needed is interactive alarm monitoring. Although interoperability is the problem confronting DIY, professional service companies—such as Vivint, which is partnering with Best Buy; and MONI, which has a kiosk in a Dallas airport—are good examples of evolving business models in which experts meet with customers to familiarize them with the technology and also to set reasonable expectations.
One important issue in the smart home is privacy protection and security. This is a growing expectation from consumers, especially as more companies look to use data obtained from IoT devices to make a profit. Today these concerns relate mostly to video cameras and voice assistants, which is why many cameras today have physical shields that cover lenses when not in use. Despite these concerns, Amazon recently launched Echo Look, likely to be used in closets—a bold move.
Privacy concerns will soon shift to routers, which will monitor devices on the network for abnormal data traffic. Monitoring software will be able to shut down devices deemed to be acting out of the ordinary, in order to help prevent botnet attacks and malicious activity. Apple has taken a lot of criticism for being slow in the smart home space, but its focus on security is unique and will likely help to win share during the next 12 months, especially as the company launches its rumored smart speaker.
Lastly, machine learning is becoming increasingly important for mass adoption of the smart home. This is because machine learning combines various benefits, such as ease of use, killer use cases and—most importantly—making automation dead simple. Many of the leading service providers today already use basic machine learning to help automatically set scenarios and complete tasks—without the user having to manually create a scenario or even needing to know about how to create one. Smart speakers are even starting to implement alerts-based notifications, building on what the ecosystem knows about the user, in order to drive an action on the user’s behalf. The new Nest camera, for instance, boasts the ability to track objects across a field-of-view and to identify friends from foe. Although this is essentially the same as video analytics—which has been around for decades—applying the features saves time and allows consumers greater latitude to do the things they love, rather than endure watching hours of video clips in anticipation of something out of the ordinary to happen.
Clearly, the path to mass adoption of the smart home is to make the “smart” in the home simple and non-disruptive to normal routine. This will include an effective pricing structure, as well as streamlined machine learning for the seamless automation of many household tasks. Furthermore, combining machine learning with
in-home displays or voice assistants will allow for a more enjoyable experience for family or friends to engage the smart home—rather than resorting to a cold, purely mechanical approach relying on individual apps that sit behind passcodes.
For more information on this and related topics, visit our Smart Home Intelligence Service.
Blake Kozak is Principal Analyst, Security Technology, within the IHS Technology Group at IHS Markit
Posted 28 June 2017