Why smart home hardware is a difficult business

My colleague Kevin Tofel rightly points out that it’s unlikely Amazon would pull the Alexa integration because it would cause the hundreds of other companies that rely on Alexa to question Amazon’s commitment. It would also do incredible damage to the Alexa brand in the eyes of consumers. Why would anyone buy an Alexa-enabled product if that functionality might one day disappear?

And yet, these things do happen. One only has to look at the disappearance of Nest gear from Amazon’s web site or the impossibility of finding Apple products at Amazon. A solution to this potential setback for businesses and consumers would be to have a data schema for all devices, an open standard. In this case, the standard could articulate what the device does along with some basic functions. With this mapped out, anyone could come along and build a voice assistant without having to spend time also developing device profiles.

So if Alexa pulled out, Sonos could at least introduce another digital assistant to the mix without having to embark on crazy partnerships and a year-long development process. And for consumers, the Sonos risk related to Alexa is a reminder that these devices don’t have hardened functionality, and it’s completely possible that the reason you purchased the product might be here one day and gone the next.

While the risk of Amazon cutting off Sonos is unlikely, the Arlo S-1 filing has a more chilling risk. Arlo makes connected cameras, lights, and security products, and hosts the data from these products on Amazon’s cloud. It notes that thanks to Amazon buying Blink and Ring, it is now an Amazon competitor, which may pose risks later on. Several retailers are trying to reduce their reliance on Amazon’s cloud, as have logistics companies. As Amazon wades more deeply into the smart home it may make sense for rivals to take their business elsewhere — or at least make sure they are ready to do so.

Amazon is also a huge sales channel for Arlo, with 16% of its 2017 revenue coming from Amazon. That means any fight with the Seattle-based company that results in Amazon no longer stocking its gear could cut out a significant chunk of revenue for Arlo.

Outside of Amazon and the business impact, the Arlo S-1 gave me pause as a consumer. It discusses engagement with its app and focuses on how similar Arlo app engagement is to social media engagement. While this might be a way to confuse investors by making Arlo seem as sexy as a social media stock, it struck me that any consumer device company focused on app engagement was really trying to be in another business.

For Arlo, that business is likely services, specifically enticing users to buy cloud storage and whatever other new service offers it can come up with. But that desire to get users to engage more deeply, combined with data that is compiled from our most intimate moments, is what gave me pause. I don’t like the psychological tricks that Facebook and other online platforms perform to get people to continue logging in, and I am not at all not excited to see the maker of a product I pay a decent chunk of change for embrace the lure of data and engagement.

There’s a fine line that connected device makers need to draw between providing services users will like based on data analysis, and luring users back again for silly reasons by manipulating them more effectively based on data analytics. With almost all of my smart home devices, I spend very little or no time in the app. Arlo’s cameras are somewhat of an exception, although I only turn to the app when I get an alert or am trying to figure out how something might have happened.

Sonos spends less time discussing data analysis, even though understanding music and optimizing its hardware for different spaces and genres is undoubtedly a big aspect of its business. In its registration filing, the more concerning elements were the haphazard revenue growth over its history and statements that it needs to continue to launch new features and products to retain both that growth and its market share.

Both Arlo and Sonos give the impression of companies doing something fundamentally challenging (building a physical product that requires a complex interaction among hardware, software, and cloud-based services) and then layering on the challenge of doing that well through multiple product launches a year.

On the financial side, each product reported significantly different gross margins. Arlo’s were around 25% in 2017 while for its fiscal 2017, Sonos’ gross margins came in at 46%. Gross margins are simply what it costs to physically produce a good divided by its sales. Higher is better, because it gives a company more room to maneuver. Sonos’ gross margins may be good, but it spends 27% of its revenue on sales and marketing. While that number seems to be stabilizing, it’s still a lot. Consumer hardware is tough.

The filings are full of useful bits of information for anyone focused on the connected home. One thing I wish I could find is a sense of the ongoing cost of supporting the existing user base. I’ll likely have to wait for the companies to go public and then see what I can learn over time.