Here’s what I think … in a year or two, saying you’re starting a company in “the Internet of Things” will be like saying you’re starting a “technology” company today. Technology has become so cheap and so easy to integrate, that within ten years we should expect Internet connectivity to be incorporated into most products and services.
Even plants. Creative new startups, like the Click&Grow units shown above in my pantry at home, are building consumer hardware products that will only get better with Internet connectivity. They use electronics to precisely water small plants based on environmental conditions inside my home. Is Click&Grow an Internet company? A technology company? An ecommerce company?
Expect to see Internet connectivity soon in anything that plugs into the wall or takes batteries. The winning hardware vendors will use IoT technology to collect data on how consumers use their products, to learn how they get frustrated with their hardware, and then to use that data to rapidly iterate their real-world hardware - whether with devices for growing plants, playing music, adjusting temperature or controlling lighting.
Often it’s better to outsource technology than try to do it yourself. United’s in-house WiFi deployment has been a disaster. I think it’s only worked on 20% of the flights where they claim to have it installed. (my experience on Virgin and Delta with GoGo has been closer to 90-95%)
Yesterday I flew back to SFO from Denver. I was excited to see the “WiFi onboard” logo, but not surprised when I got this result over and over:
Later in the trip, I asked the flight attendant if she was headed home. “Nope. Headed to San Diego. About to get on another Airbus where the WiFi doesn’t work.”
Funny thing was, I didn’t mention WiFi to her. At all. What it means is that United is disappointing a lot of their customers. Likely the ones who pay the most to fly.
What do you mean, I asked, curious. “I can’t remember the last time WiFi worked on one of these planes.”
When I got back to my seat, I could see that the guy next to me was trying to log on as well. He said that he picked this flight because it showed WiFi in the schedule. He also said that he wished he picked Virgin instead.
There’s really no excuse for this if you care about your product. I’m guessing that someone at United is thrilled to be able to say that “x% of our planes have WiFi” to entice customers, but the sad reality is that the only WiFi that seems to work on United is on their PS transcontinental flights which are powered by GoGo. Yes, everyone likes to complain about airlines. But why have other airlines gotten WiFi right?
I wonder that this does to the NPS for United. It’s one thing when you don’t have a service - it’s another when you promise a service and fail to deliver.
Will wearables happen? Yes, of course they will happen.
But that’s a different question than whether it’s a great place to start a company today. It’s a crowded market, and although early pioneers like FitBit and Pebble have proven large latent demand, the consumer value proposition is still a bit uncertain.
In a crowded market with uncertain consumer value propositions, the companies with the biggest marketing budgets win. Think Samsung, Google, and Apple. Most of the startups I see in the space have modest incremental improvements. I’m aware of at least 50 startups with these modest, incremental improvements. To win in this market, you need to have a 10x improvement on a dimension that matters - something like monitoring blood chemistry in real time, non-invasively. That’s the type of wearable company we’re looking for.
Earlier today, I gave the following talk at Montreal’s Startup Festival on how to think about the future of technology.
Investors and entrepreneurs can make a lot of money when predicting the future – by being right when others think you’re wrong. There’s been a lot of talk in Silicon Valley lately about “where’s the real innovation going to be?” which has been summed up by (paraphrasing Peter Thiel) the comment “…where are the flying cars?”
The reality is that most of the things the pundits predicted (summed up in science fiction franchises like Star Trek) about humanity’s future require incomprehensibly large amounts of energy. They’re simply not going to happen anytime soon. At the same time, without us really noticing, the key plot elements of “enchanted objects” have emerged around us - even physical invisibility is closer at hand than we’d like to admit.
For the next 20-30 years, massive, predictable and relentless reduction in energy requirements in semiconductors will lead to the Internet creeping into everything. Moore’s Law today is about a 50% reduction in energy requirements every 18 months as much as it was about increases in processing power and memory density.
As much as I’d love to fly to the moon and Mars, moving physical objects still requires enormous amounts of energy. With the exception of atomic energy, energy improvements only increase at the rate of 2-5% each year, which pales in comparison to Moore’s Law. It’s going to be a lot cheaper (and faster and more comfortable) to fly to the moon in the “magic mirror” simulator of a VR headset like the Oculus Rift than spaceships, and to many consumers, the VR experience in 30 years may even feel “more real” than the real thing.
Where’s the flying car? Forget the flying car. The enchanted self-driving car is right around the corner, is just as awesome, and early prototypes are already around you on the road today.
RelayRides is really hitting its stride - it’s no accident that the company recently closed on a $25 million round.
A few years ago, I bought a 2001 BMW Z3 just to rent it out on RelayRides. It’s done phenomenally well. If I wanted to, the car could be rented out all of the time - I’ve seen a huge amount of inbound demand.
Great p2p marketplaces constantly add value. Google “BMW Z3 rental San Francisco” and see what comes up at the top of natural search. (hint: my car) RelayRides helps market my car - via both organic search in Google and direct search on RelayRides.com. They vet renters for me, making sure they’re safe to rent my car. They take care of deposits and provide escrow for payments. They help me properly price my rental rates. Once, my car came back with front-end damage. Minimal paperwork was required - all I had to do was forward pictures of the damage using my iPhone, take the car to a body shop of my choice, and forward the estimate via email to RelayRides. A check showed up a week later. No need for me to wait on the phone for hours to talk to an insurance company or to the renter about the damage.
Great p2p marketplaces also require regular tuning. A recent piece in Fast Company talks about why RelayRides exited hourly rentals. Hourly is not a great business - either for renters (thanks to Flywheel, Lyft, and Uber alternatives) or for owners (higher claims rates and lower revenues per rental). And now that RelayRides has achieved significant early liquidity and scale (4x+ their closest competitor on all dimensions), the company has statistically relevant claims history which is driving down the cost of insurance and leading to attractive gross margins.
I’m excited to see how big this company can get. The company has grown organically 3x year/year for the last several years. The team got to this point on a relative shoestring, and now they have a growth round to let them really show what the company is capable of.
Last week we announced the formation of Shasta IV, a $300 million early-stage venture capital fund. For ten years our team has pursued the same strategy – focus on Series A investing, with some Series B and seed investments as well. In a world of mega-funds and seed funds, at Shasta we’ve decided to take a boutique strategy with 8-10 core investments each year, in targeted sub-sectors of consumer and enterprise technology businesses.
This means that there are a lot of things we don’t do. We don’t feel a need to invest in every hot trend that comes along – instead, we focus on areas where we think the opportunities are bigger than others realize, and in areas that the partners find particularly interesting on a personal basis. For me, that meant investing in the Internet of Things long before it became a household phrase (like Nest Labs, for example), and in connected hardware since I’m a mechanical engineer and love startups that build physical devices. This means that I can better help the companies we back in sectors I deeply understand, rather than trying to become an expert in everything.
I’m very thankful for the support Shasta has received from our earliest investors and the entrepreneurs we’ve backed over ten years. There were only a few new limited partners that joined us in the latest fund; our overwhelming support came from investors in our prior funds. These investors talked to many of the entrepreneurs we’ve backed, and we appreciate the extremely positive feedback these entrepreneurs shared with our investors. Our investors also told us that we have a high rate of “repeat entrepreneurs” – founders we’ve backed in the past coming back to us again – a statistic of which we’re particularly proud. Of course, we will continue to enthusiastically build new relationships with young first-time entrepreneurs. Every day, I see interesting new teams and ideas, and I’m sure that the most exciting new companies will head in directions we haven’t even thought of yet.
"How likely is it that you would recommend [insert product name here] to a friend or colleague?"
This simple question, answered on a 0-10 scale (0 means not at all likely, and 10 means extremely likely) tells a company everything they need to know about whether their product will benefit from strong demand or suffer from high costs of customer acquisition. You can see the overall Net Promoter Score (NPS) methodology here.
Word-of-mouth is life or death for a hardware startup. The marginal cost of producing additional units is high, and unlike software, hardware companies (even today) need to invest in inventory. Once you go into volume production, your product better fly off the shelves. And only strong word-of-mouth, massively amplified by social media, will achieve that. The best predictor of strong word-of-mouth is a Net Promoter Score north of 70.
So what does that mean for a hardware designer? It means that every design decision, every single one, should be made with your NPS score in mind. Ask yourself each time: ”will this decision increase or decrease the likelihood of customer recommendation?”
Should I add this feature? Should I use this material? A bigger, smaller, brighter or dimmer display? Will it cost too much? What color should it be? Knowing your customer allows you to understand how to trade off these items - and if you don’t really know your target customer, you won’t be able to make the right tradeoffs. The trick, of course, is that you can’t measure NPS until the product is done. But if you really know your target customer, you can climb into their shoes and design your product appropriately.
After Google’s $3.2 billion purchase of Nest Labs, a lot of people are trying to predict what’s next in the home. During the last few months, I’ve met with dozens of startups trying to create platforms for the Internet of Things - both in the home, and in the enterprise for industrial applications.
I think a lot of these startups don’t understand where IoT technology applications are headed. I believe in the value of “silos” - individual use cases that work well on a stand-alone basis, but don’t have a lot of interaction with other “silos”. In the home, I’ve installed Sonos in most of the rooms in my house. I really don’t see the utility in connecting my Sonos equipment to my Nest equipment. Or my lighting. Or my locks. I already have the convergence layer I need - my smartphone.
Yes, it would be nice to have hot, fresh coffee ready for me as soon as I wake up in the morning. But the challenge is not in automatically brewing the coffee before I get out of bed, or using motion sensors to detect me stirring in my sleep. I still have to change the filter, wash the carafe, and fill it with beans.
The real value in connecting the coffeemaker to the Internet will be for the manufacturer. Imagine a world where any device that’s powered and plugged into the wall can capture usage data for manufacturers. What if Keurig or Breville knew exactly how often their coffeemakers were used? Which features were used? Does usage of the device fall off, or remain constant with time?
Imagine an Internet-connected refrigerator that monitors how often its water dispenser was used, and then sends owners fresh filters when appropriate. Imagine lamps that can tell when the light bulbs burn out, and automatically sends replacements. Imagine coffeemakers that measure usage and work with Amazon or Google to replenish your filters and beans.
This is the future of smart devices and appliances. When parts are cheap enough, every device will “phone home” to their manufacturers (via BLE, LTE, or WiFi), and savvy companies will use this data to quickly discover features and capabilities their users really want - and build better differentiated products more quickly as a result. They’ll also be able to learn how things break, and improve product reliability faster as well.
Just like today’s top performing web properties, A/B testing is coming soon to appliances and anything else that’s plugged into the wall. Just like when you take a Sonos speaker out of the sealed bag that has a “user agreement” attached to it, most of our electronic devices and appliances will come with Terms of Service as well.
As a mechanical engineer, I believe in first principles.
The first principle of building a consumer-facing startup is to recognize that consumers are rational. They may deny it to themselves, but people always look out for their self-interests. Their self-interests may be driven by helping others, or driven by vanity and selfish goals, but people always seek to feel better about themselves. That satisfaction can be realized in many ways – some of which are self-serving; some of which are selfless.
When I think about consumers, I seek understanding and counsel startups to do the same. Suspend judgment. Seek understanding. Think about the state of mind of your end-users. Did they just wake up? Are they driving home? Did they just get bad news? Did someone in their family just die? Did they just get great news? Did they get a haircut they really like? Did they just get a nice note from a friend? Did a cute person just randomly smile at them? On and on and on and on, our daily consciousness is overwhelmed with emotional interactions and feelings.
Every consumer in the world is flooded with messages that compete for their attention. How do you differentiate and get above that noise? How do you convey your value proposition in less than a minute? In less than 10 seconds? Don’t judge consumers for their short attention spans – accept it – and figure out how to make their lives better with a product or service, and then message that efficiently and effectively.
That’s the “mystery of life” in consumer-facing startups. If this bothers you, don’t bother building or investing in a consumer company. But if you love this challenge, and accept people as they are, jump into consumer-facing businesses.