The team at Panic decided to discontinue their Status Board product and wrote a brief post describing their decision. Sunsetting a product is an emotionally difficult decision, and I empathize with them. The TL;DR on their post-mortem is simply there wasn’t enough demand for their product to justify the cost of continuing to support it. The part of their post that caught my eye, which prompted me to write here was:
Finally, in the pro/corporate universe, we were simply on the wrong end of the overall “want a status board” budget: companies would buy a $3,000 display for our $10 app. Hmm, maybe we should’ve gone into production on LCD displays instead…
The final sentence about producing LCDs is tongue-in-cheek, so I don’t want to overanalyze this specific example, but consumer electronics (CE) manufacturers’ envy of software companies is significant. They see software folks with their negligible COGS, their direct relationship with customers, their rapid iteration cycles, their recurring revenue business models, and they are deeply envious. CE companies know they sell commodity products that are easily swapped out for each other, cleanly abstracted away from other layers of the stack by standardized I/O ports. They know that, at the end of the day, they sell fancy glass, easily measured against other fancy glass in itemized feature comparison charts. And so, they produce software to bundle with their glass in hopes that their glass will be interpreted as differentiated or unique in less tangible, less commodity way. Every TV sold this decade has been advertised as “smart” and bundled with mostly unusable software, all built in-house by inferior software talent (when compared to products from companies that only build software for a living) with the goal of directing consumers’ spend on content. They think Apple’s 30% app store tax is pretty sweet, especially because it’s nearly 100% gross profit, and they want some too.
If the software guys are envious of the hardware guys’ top-line retail price points, and the hardware guys are envious of the software guys’… everything else… then where is Goldilocks’s happy “just right” bowl of porridge to be found? Who are the companies that sell solutions that are bought as atoms with atom-like price points, but have all the flexible benefits that come with great software and the continuous relationship they maintain with their customers as the product improves over time?
Apple is the most obvious answer. Many ink-pixels have been spilled over their success in integrating software and hardware such that I don’t need to reproduce a cruddy facsimile here.
Fitbit certainly pulled off this jujitsu move too. People hem and haw over the purchase decision of a $2 fitness app for their phone (and ultimately all fitness apps that survived the app store gauntlet went free and found other paths into consumers’ wallets), but a $60 fitness band with software to help you compete in fitness competitions with your co-workers or friends feels like an affordable, snackable, giftable purchase price.
The hardware/software combination was the magic of Dropcam too, which allowed them to pair mostly commodity webcam hardware with a software subscription service. I don’t know Dropcam’s subscription retention, but I’d bet a nickel that, when they were an independent company, they were on track to make more revenue from software than from hardware in their customer lifetime value equation. Their software was wrapped in injection-molded plastic, sold as atoms, and vertically integrated as the only option to make your new camera really sing after purchase, for a monthly recurring price. My partner Santo taught me the phrase “tin-wrapped software” awhile back, and it made a lasting impression.
This post is much more of an observation than a recommendation. As an end-consumer, I much prefer to buy commodity hardware that is cleanly abstracted from its software stack, which allows me to make independent software choices from other vendors. I’d be sad if every company that sits somewhere close to the hardware/software border took to heart the examples I cite in this post because then all my choices in the future will be vertically locked solutions that allow someone like myself no opportunity for customization. But, I’m a geek. I take pleasure in tinkering. Another aspect of the hardware/software combination magic is the low cognitive overhead for setup. No pairing devices, no figuring out which input am I connected to, no compatibility issues, just one purchase decision.
My gut says that vertical integration between hardware and software is on a pendulum. Many PC solutions in the 80s were deeply vertically integrated to the point that word processing documents from one brand couldn’t be properly read and edited another brand. Today, that experience sounds absurd, because we have benefitted from decades of work on cross-compatibility between computing platforms. For embedding computing and IoT, we are seeing a pendulum swing back towards vertical integration, which will bring with it incompatibility issues, primarily because combining hardware and software well can deliver a magical experience and a revenue premium. But, we could find that the 2020s IoT landscape rhymes with the 1980s PC compatibility wars. I look forward to seeing how it all evolves.