When a friend forwarded this tweet from Paul Graham, it hit close to home:
Startups are subject to something like infant mortality: before they’re established, one thing going wrong can kill the company. Hardware companies seem to be subject to infant mortality their whole lives.
I think the reason is that the evolution of the product is so discontinuous. The company has to keep shipping, and customers to keep buying, new products. Which in practice is like relaunching the company each time.
I don’t know if there is an answer to this, but if there were a way for hardware companies to evolve more the way software companies do, they’d be a lot more resilient.
Looking back on our startup journey at Minut, I remember several moments when we could have died. However, surviving several near misses we learned to tackle these challenges and have become more resilient over time. While there will never be one fully exhaustive answer, here are some of the lessons we learned over the years:
Subscription revenue is the only revenue that counts
While you can sell hardware with a margin and make important early revenue, it’s not a sustainable business model for a company that requires both software and hardware. You can’t cover an indefinite commitment with a finite amount of money.
Many hardware companies don’t consider subscriptions early enough. While it can be hard to command a subscription from the start (if you can, you might have waited too long to launch), it needs to be in the plan from the beginning. Look for markets where paying subscriptions is the norm rather than markets that operate on a one-time sale model.
Set high margins and earn them over time
It’s tempting to set low prices for hardware to attract customers, but in the beginning you should do the opposite. Margins allow for mistakes to be rectified. A missed deadline might mean you have to opt for freight by air rather than boat. You might have to scrap components or buy them expensively in a supply crunch. Surprises are seldom positive, and you don’t want to use your venture capital to pay for them.
Healthy margins can also be used to cover marketing costs while you learn what kind of messaging works and what channels you can sell through. If that wasn’t enough reason, starting with relatively high prices will help you avoid another common mistake, selling too much at launch.
This might seem counterintuitive — why wouldn’t you want great success out of the gate? The reason is that you will inevitably make mistakes with your early launches, and the bigger the launch, the bigger the blow. There are plenty of companies who achieved amazing crowdfunding success and then failed to deliver even the first units. Startups tend to chase growth at all costs, but for hardware startups in the first few years there is such a thing as too much of a good thing.
Choose your customers wisely
Customers with long decision cycles or many stakeholders are bad initial targets for hardware startups. Start at the lower end, with smaller businesses or even consumers. You can move upstream or sideways later.
Each iteration toward product-market fit requires feedback from customers. You can get fast feedback from fake ads, landing pages and interviews, but no feedback is more valuable than that coming from someone paying money for your product and relying on it daily. That’s why you need to pick a target customer that can make fast decisions so you can get to market quickly.
Having customers capable of making fast decisions counteracts the inevitably slower speed of developing hardware. After all, what really matters is not just how quickly you can release new features, but how fast you can complete a full cycle of customer feedback and product iteration.
Avoid the launch trap
Airbnb famously launched seven times. If no one notices the first few times, why not try again? For a software startup it can seem like a no-brainer, but for hardware startups each launch is a potential trap. The revenue generated by a crowdfunding campaign or pre-launch of a shiny new product can be addictive, but it forces you to deliver on promises even if the result isn’t what you hoped for.
Worse still, since it’s easier to sell new features to existing customers via crowdfunding than winning customers in the regular market, you can become reliant on selling hit after hit to crowdfunding enthusiasts. Learnings from this group don’t transfer well to the broader market. Minut fell in this trap.
Instead, iterate on the core functionality of the product in order to reach the customers you didn’t convince the first time. Focus on the real pain and not the feature requests. Your product should ideally never go through a revolutionary change — much like the iPhone gets better each year but never startlingly so.
Decouple the customer experience from the hardware
Your product is not the hardware it requires to run. While hardware iterations take time, you can iterate on customer experience just as fast as software startups do. The whole system will have higher complexity, but you can architect it so that each part can be improved independently. Treat interfaces as if they were external (one day they might be) and make hardware versatile enough that it can support multiple product iterations.
At Minut, our hardware platform is versatile enough that it’s being used as a home alarm, a sensor for office management, in elderly care and (of course) with Airbnb in hospitality. Versatility, platform thinking and flexibility are imperative on the journey to product-market fit. There is plenty of time to focus narrowly once you’ve found it.
Innovate on the product, not in the supply chain
When you hear hardware startup founders talk about close encounters with bankruptcy, their stories frequently involve the supply chain. While most startup decisions are reversible and bear modest consequences, almost every supply chain decision can be fatal. Whether it’s choosing a supplier that turns out to promise a bit too much, or a fancy new chipset that didn’t really live up to the data sheet, you can lose a lot of time and money if you’re not paying attention.
The good news is that you don’t need to outperform. Winning at supply chain for startups simply means not losing, and best practices are well known, so pay attention and stick to them. Don’t be tempted to shave off a few dollars by using unknown components. Pick suppliers that are the right match for you. Test every feature of every unit in manufacturing and never ever try to rush to production without a proper process.
When running a business, you need to focus on making a few bets that have substantial upside with relatively low downside risk. Innovation in supply chain management usually gives you the inverse: limited upside with serious risk at every turn.
In short, Paul Graham is right — there’s a staggering amount of ways a hardware startup can die. Fortunately, the most common mistakes are well-known and if you learn about them from others, there are also numerous ways to avoid the pitfalls.
If you’re a hardware startup founder and want to connect, please reach out at nils@minut.com!