4 Billion-Dollar Opportunities for Hardware as a Service
In a previous post, I already outlined why we at Point Nine Capital believe that there’s a new era for hardware startups.
As a quick reminder, in our thesis for Hardware-as-a-Service startups, we are looking for:
- Subscription/recurring business models
- Online direct distribution channels - content marketing, SEM, etc.
- Products that are almost software - where most of the intelligence is on the software side, not the hardware device.
Here, I want to share some areas where we think there might be great opportunities to be captured:
Sensing the world
With the cost of connecting devices going down, it’s becoming more viable to collect data from the physical world that wasn’t available before.
That has two very interesting implications:
There’s plenty of areas where collecting more data can have a significant impact on the performance of a business.
Just think about supply chain and manufacturing optimization, analyzing return of investment in marketing, different insurance products based on customers’ behaviors, agriculture optimization, farming, etc.
- Euclid Analytics was (probably) the pioneer on that trend. They provide something close to a Google Analytics solution for your physical retail location.
- Metromile provides insurance for drivers based on their driving behavior.
- Enevo uses wireless sensors to measure and forecast the fill-level of waste containers in cities and generates optimized collection plans.
- Automile (where we are investors) gets data from your car to allow fleet managers to understand usage patterns
Companies can’t backup the physical world
While in the digital world we are very much used to backup our data in case a disaster happens, that’s impossible in most cases in the physical world:
(1) We can copy digital data into a different location — hopefully safer! But in the physical world we might struggle to collect the data without (connected) sensors.
(2) We can recover digital data and “go back in time”, but it’s a lot harder to do so in the physical world.
Let’s try to explain it with an example: developers release software continuously, but if there’s a bug in the latest release, they can go back to a previous one without that bug. That allows them to iterate a lot faster in their development process because they are less afraid of the mistakes they can do — they can always go back in time!
So far, I haven’t seen any true example of that idea.
In my opinion, two examples that come close to that are:
- The nest thermostat (1) keeps recording data about temperatures and (2) then sets the thermostat to different settings at different times.
- Angelcam.com allows you to connect your camera to the cloud and record videos (1) — it doesn’t allow you to travel in time to that moment, but at least you can rewind your videos to that time ;)
Please comment if you have some other ideas.
Interacting with the world
If you think about the great revolutions in communication — telephony, internet, smart phones… — all of them happened through hardware devices. They enabled us to communicate — first via voice and later on via data.
Still today, most of the devices we operate on a daily basis are pretty “dumb”.
Even the smart devices tend to be limited to sensors collecting data, but they are unable to trigger actions in the real world.
Connected devices are opening an era of sensor to smartphone communication — which I believe, in the future will become sensor to machine communication.
On top of that, as a developer you are lucky to have APIs. APIs make it easy for different pieces of software to intercommunicate among them, exchange information, trigger actions, etc.
Unfortunately the real world doesn’t come with APIs by default ;)
But that’s already changing.
The most well known cases are in the smart homes and offices’ space:
- Nest’s thermostat learns from your activities to make heating control smarter.
- KISI (where we are investors) triggers your office doors to open or not based on the permissions stored in your phone.
What’s even more exciting is that this trend doesn’t end in the smart office/home. I think it actually might have significantly larger implications in other sectors with heavy machinery or that are very intensive in human labor like agriculture, farming, manufacturing, construction, etc.
“Softwarization” of hardware
Customers are increasingly willing to buy hardware that behaves a lot more like software —that means: easy to manage on remote, mobile-first approach, online backups, APIs, sold as a subscription service rather than owning it, self-service distribution and on-boarding of users, etc.
The first cases of that have been around connected/networking devices like
- Sonos connects your HiFi system to an app
- Meraki which provides cloud connected WiFi access points
- Dropcam which updates security cameras to the cloud era
In this specific case, it also looks like open source hardware might also play a very relevant role.
But not only that, in some other cases, customers are not willing to buy the hardware at all and are preferring to rent it. The obvious leader on that is Amazon Web Services, but I’m pretty sure there’re other areas where that model can be applied too!
That deserves a blog post on its own ☺ but obviously there’s billion dollar opportunities here too
UPDATE: Here you have the drones post