14 Lessons Learned Investing in B2B Hardware-as-a-Service Startups
Even with some of the most successful companies in the world being hardware companies — like Apple, Bloomberg and Cisco — investing in hardware startups has been out of fashion in “silicon” valley. ;)
From a software investor perspective, hardware is particularly hard because hardware companies:
- can be riskier: it’s slower and harder (no pun intended!) to iterate until you find product-market fit.
- can be less capital efficient: it’s more expensive to develop, manufacture, and distribute hardware products.
- can have limited upside: it’s challenging to create long-term moats for hardware startups.
But I believe that this “mantra” can be questioned today as Software is also eating hardware. The best example of this is AWS, which is now a $15B+ business. Can we consider it as the most successful HaaS product launch of the decade?
At Point Nine, we have been, and continue to be, very bullish about the opportunities for B2B Hardware-as-a-Service (HaaS). During the past few years, we’ve looked at many companies in the space, and also invested in several such as Automile, Kisi and another one that hasn’t been announced yet — stay tuned for more info!
Working together with founders we learned a ton about the specificities of the HaaS business model. In this post, I’d like to summarize some observations that can hopefully help other founders:
Lesson #1: Hardware Development is Still Pretty Damn Hard
If you combine the facts that:
(a) the average early stage HaaS company has a bit more than a year of runway after raising its seed round.
(b) the iteration cycles for hardware startups are still in months (and delays are common).
You quickly understand that there’s not a lot of room for mistakes.
Luckily, the “softwarization of hardware” trend that I mentioned above helps to shorten the iteration cycles. On top of that, today, there’s also a broader range of outsourcing partners that are happy to work with startups.
We are also lucky to see more teams with experience in dealing with this kind of projects, and this experience is very valuable to avoid mistakes that can be expensive in the life of an early stage startup.
2. It’s Still Hardware — Save +20% of the Money Raised for Inventory
In the introduction, I explained that hardware companies have costs, like manufacturing and shipping, that software companies don’t.
Luckily, today some external providers offer startup friendlier terms, like smaller production batches, and you can also leverage the existing eCommerce infrastructure when it comes to shipping or storage. Nevertheless, you should assume that more than 20% of your seed funding will go into inventory costs. Unfortunately, debt financing only tends to be available at a later stage, but hopefully, it will change.
3. Interacting with The Physical World Allows Unique Value Propositions…
Pure software products can do amazing things, but the hardware interface where they run (phones, smart speakers, and desktop interfaces) create constraints on the kind of problems that can be addressed, and of solutions that can be built.
Dedicated hardware products enable founders to remove some of these constraints. Many problems can be solved only with specific hardware. Broadly speaking I see two high-level categories:
a) Sensing the world
Phones can be used to gather data using their camera, GPS, microphone, etc., but specific hardware sensors can enable broader use cases (i.e. temperature, humidity, etc.) with cheaper costs, different setups (at a fixed location, glued to an object, etc.), and higher speed or better accuracy, among other benefits.
b) Controlling the world
Phones have become a “remote control for parts of our lives”, but most physical objects are not connected, so we still need hardware devices to interact with the real world.
4. … But When Dealing With the Physical World, “Don’t-Take-Anything-For-Granted™”
In the software world, we take for granted many aspects that define the end-user experience. For example, if you provide a SaaS product, you take for granted that the user will have internet :)
In a HaaS business, a great product is worth nothing if the Wifi doesn’t work or if somebody unplugs it (real cases!). And it’s not only about the WiFi or the battery power, the fragmentation in terms of devices (sometimes you need to be compatible with many different devices in different settings) can also be a pain to deal with. This is why bullet-proof testing and on-boarding processes are essential; not just nice to have.
5. It’s Hard = Limited Startup Competition …
For all the previous reasons, it’s not easy at all to start a Hardware-as-a-Service business. As a consequence, you might not face a lot of competition at the early stage.
If you’ve managed to build something that “just works”, that’s already a significant achievement! First, please call us. Second, don’t be surprised if you’re the only new player in your market.
6. It’s Hard = … and Sleepy Incumbents
Most of the large hardware incumbents aren’t software-first companies. When they speak with customers, they insist on their hardware product and forget about how powerful software can be. In this situation, you can win with a “consumerization” playbook: build a slick software that caters to the needs of the end-user. Value the software. Commoditise the hardware.
7. Potential Customers Locked Into Existing Long-term Contracts
Unfortunately, not everything is so easy.
Many of these incumbents are in oligopolistic situations: think about routers, telematics, access controls, video and sound equipment, printers, etc. You can expect them to use every single trick to extract top dollars from their customers and to lock them in for a long time.
Even if you manage to build a 10x better product, they can still lock you out of potential customers with multi-year contracts. Be patient. If you try to grow faster than the market allows, you may just increase your burn unnecessarily :)
8. There Are Budgets! Move them from CapEx to OpEx
The previous issue can make your life difficult when you want to win customers. But at the same time, selling a hardware product has its unique benefits:
The good news is that in most cases there are already budgets in place if you’re replacing an existing physical solution. And often those are significant budgets for capital expenditures (CapEx in the finance lingo), so a good argument is to transform these into recurring subscriptions which are then considered as operating expenditures (OpEx). From CapEx to OpEx.
9. Software-like Margins
Even with the costs inherent to hardware products, we have seen that HaaS companies can achieve software-like margins.
When you look at the revenue profile of a customer, you’ll see that, generally, the hardware, marketing, and delivery costs can be recovered in the first year. After that, the revenues generated by the subscription have typical software margin.
10. Online and Self-service Works!
The demand generation techniques used in the SaaS industry can work well in the HaaS world. You can generate leads through content, SEO, paid advertising etc. And if your product is self-service enough, prospects can order it directly from your website, and you can even offer trial periods.
11. You Can See Higher Conversion Rates …
Today, testing a new software product as a customer is extremely easy. You can go online and try a product without even giving any credential nor credit card details. This race to remove friction in the onboarding process is probably one of the main reasons why the conversion rate from “trial to pay” is low and even getting lower in the SaaS world.
In the HaaS world, this is not the case yet, and if you have a great product you could be pleasantly surprised by conversion rates that are actually pretty high!
Lesson #12: … and Lower Churn …
Not only can conversion rates be significantly higher, but churn rates can also be significantly lower.
I don’t have a great explanation about that, only an hypothesis: I guess that, compared with software companies, the friction to set up a physical device can also make it stickier. Customers don’t make an effort to look around for another solution (maybe out of laziness?).
13. … and a Pricing that Scales
The best friend of a SaaS business is negative churn.
But that’s not an easy quest! To get there, your existing customers must pay you more and more over time to compensate for the lost accounts. The main lever for this is a pricing strategy where one variable correlates with the value delivered, as well as with the size of the customer. In the SaaS world, the common way of achieving that is by using a seat-based pricing — but that doesn’t always work!
In the hardware world, that can also be achieved, and in some cases, it’s even easier. Because you have the device as a natural variable for adoption. For example, Automile charges customers per device with different pricing tiers depending on the features available. Kisi achieves the same goal through pricing per door and per user.
14. All of this can be Defensible :)
Fitbit and Gopro have shown how challenging it can be to build moats for a consumer electronics business.
In the B2B HaaS space, we have mentioned that companies can enjoy a lower churn because of “customer laziness”. But that’s not the only reason ;) Other traditional sources of defensibility of SaaS companies also apply — such as brands, network effects through platforms, deeper tech, etc.
Are you an entrepreneur building a HaaS company?
I would love to meet you! Please follow me at @DecodingVC
