Hardware-as-a-Service | Hacker News Reader

Hardware-as-a-Service

diff.substack.com

40 pointsalkhidr34 comments

Hardware is hard, and, ironically, crowdfunding services make it even harder. I always feel bad for Hardware startups who go through crowdfunding services like Kickstarter. They almost always underestimate the cost of development of the final product, while at the same time giving (deep) discounts on that final product to early backers. In this way, they build up a huge amount liability because they burn through the Kickstarter cash real quick.

I work on (embedded) hardware and completely agree with you, with a couple of addenda. Hardware crowdfunding is a great way to check for some product-market fit, and access early adopters; it isn't a great way to fund hardware R&D, especially for commoditized goods (where pricing is dictated by competition).

I worked with both Kickstarter and Crowd Supply, and I have to say that the latter can help you significantly.

Cashflow is certainly an issue. In our 1st KS we definitely did significant discounts. Thankfully we remained profitable, but what we didn't really considered well enough was the amount of money you need to build inventory AFTER the campaign.

Rule of thumb now: 3x your production cost. 1x for manufacturing, 1x for various expenses including discounts, 1x for next batch.

That squares pretty well with the old rule of thumb for the cost of opening a restaurant or bar: Figure out what you think it will cost, then triple it and add a third.

That's a great perspective. I actually think Kickstarter could be great, if backers really treated it as a way to fund projects they like, instead of a pre-order of a final product with an early-bird discount. That is, if you're building hardware with a target price of $100, early backers should be paying $300+ for that product (or $x no-string attached donation + small $ amount for the right to be first to pre-order to be used against the final price of the product) because otherwise, yeah, like you said, you're going to use the Kickstarter capital to finish development and have no money left over to fulfil orders.

As someone who has developed hardware for >10y now, there’s one crucial mistake a lot of startups make:

- Thinking that making hardware is profitable.

Don’t get me wrong, there’s some markets where hardware is profitable, but they are niches, and the prices might surprise you.

As a reference, check the mechanical keyboard scene. Check out how much the components cost, and at what price those keyboards are sold.

When it comes to general consumer hardware, a cursory look to hardware companies will show you that the margins are razor thin, the risk is real, the problems are many.

You don’t build hardware to make money, nowadays you make hardware because it enables you to sell software. That’s a better approach for most “hardware” companies.

This may be true, but i think keyboards are a bad example. Keyboards are generic, there's been very little innovation in a long time, everyone knows how to make a basically equivalent product. The only real distinguisher between competitors is cost. That's going to drive margins to be razor thin. But not because its hardware but because the type of product a keyboard is. You would probably see the same thing in the market for hammers, cutlerly, etc.

Mechanical keyboards, not regular keyboards.

That's a wildly malformed market too.

You have a lot of oddities in the mechanical keyboard hobby.

* For a while there was a lot of rebranding. Three or four of the most popular boards a few years ago were fundamentally the same Costar CST-104 with different logos.

* Some of the legacy players (Cherry and Unicomp) were bankrolled by sales outside the enthusiast market.

* There's a huge number of hobby projects that overlap heavily. A million "40/60% with an aluminium case and takes MX switches" kits. I suspect this offers poor economies of scale and the ability to hide large margins.

I started Framework to build an fourth option, which is to monetize the longevity of consumer hardware through a marketplace for repair parts, upgrades, and used product/part re-sale. In this way, we align incentives between the OEM, the consumer, and the environment. If a consumer is not using a product, we want to incentive them as much as possible to input it back into the marketplace where it can find another consumer (and we capture another transaction). If a consumer is using a product, eventually some part of it will wear out, break, or go out of date, and we want to enable that user to keep using the product by getting them the part and/or service they need to keep using it.

The goal ultimately is to bootstrap a large enough install base on each product category we enter to make it worthwhile for third party compatible part/product makers, refurbishers, and service providers to participate in the marketplace. The resulting ecosystem of providers makes the product and marketplace serve more consumers better, fostering network effects.

I hope you succeed, because that's what we all need!

I have a totally boring "financial engineering"-related dream every time I need to replace a major appliance: I would gladly pay $X/mo for one appliance in perpetuity that I never have to think about it again. Give it a replaceable facade and no one would ever know when it's 10+ years old.

I don't need fancy apps, or screens, I don't want the internet on my fridge... Just give me enough space for my food and keep it the right temperature! Same with the dishwasher, just make it work well and don't build everything out of plastic and I bet it would last a lot longer...

Water heaters and furnaces can be leased instead of purchased, for a few dollars a month. (But it's more expensive in the long run.)

What's different about a furnace vs a fridge or dishwasher?

Furnace longevity depends on how good the owner is at maintaining clean filters, and water heaters depend on water quality. I'd find it difficult for either of those to be offered as-a-service at a reasonable price considering the aforementioned volatility versus just selling them since they're not that expensive in the first place.

Yes but in that case it would be in the interest of the leaser to make the product be as easily maintainable as possible.

> ... it would last a lot longer ...

But they can make more money if you buy a new one every 8 years instead of every 10.

But "they" can make more money if they sell a better product than their competitors for the same price.

There's no monopoly on dishwashers and other appliances. People simply choose to save money and go with cheaper items, or they can't afford ones that come with a 10 year warranty. Or maybe they don't exist. I know miele products exist, but not everyone wants to spring for Miele or Speed Queen washers or other higher quality builds that offer longer warranties.

I know I'd rather save $1,000 or more, invest it, and take my chances with an appliance from Costco since I can get it with a 4 year warranty anyway. If it breaks after 4 years, I'll get a new one. Over the past 10 years, I haven't experienced any appliances breaking.

>I would gladly pay $X/mo for one appliance in perpetuity that I never have to think about it again.

That's a very common model. For example, our business didn't buy a printer/copier - rather it is pay-per-use and ostensibly under perpetual warranty while we're paying. If something breaks, a tech comes in and fixes it or replaces the unit. Coffee machines and commercial dishwashers are another example. All these can be leased and serviced for a monthly cost and typically don't require any up-front capital. It's very convenient but more expensive than an outright purchase.

This model works well for commercial enterprises, but it just isn't worth for consumers. It's like car leases - bad deal for consumers, but attractive due to flexibility for businesses.

That's the thing tho. Make it cheap enough to not be more expensive. Make it a quality product so there's little maintenance cost. Don't replace it every 10years and there's less overheard across the board.

> Make it cheap enough to not be more expensive

Are you expecting it to be cheaper to rent than buy in the long term? That's pretty impossible since the person you are renting from has to buy at least 1 and provide additional services. It doesnt matter how cheap you make it; x<x+1 for all x no matter how small x is.

I think there's a few possible angles here.

* Cheap and Expensive as a subjective thing. Paying a $10 per month fridge bill in perpetuity may feel less uncomfortable than the sudden expense of a $700 fridge every 8 years, even though the latter is cheaper.

* A service can take some of the bad-decision factors out of it. Many appliances are panic-bought when the previous unit fails, rather than waiting for the price to bottom out. A predictable lease price and service organization avoids that risk. It might not be as cheap as the cheapest unit on the best day, but it's competitive with a mid-line product.

* The rental service product could benefit from economies of scale and standardization that direct-to-consumer brands don't offer. Whirlpool sells a few dozen different 25-cubic-foot fridges, with constant model churn. Any given model may only have a total run of a couple thousand units. Fridge-as-a-service may only offer one, and offer it for 10 years, allowing slower amortization of engineering costs and less expensive service/spares management.

* A structured end-of-life program could bring down the cost of leasing. After five years, they swap out the old fridge, but then they can refurbish it and either return it to the lease pool or resell it in a less ad-hoc manner than listing individual units on Craigslist.

This model does not work very well for consumers. It really does raise the prices of goods and drains consumer income and forces an opportunity cost of not investing in retirement, and not paying down mortgage faster (for example). It's almost always better for consumers to lay out the capital up-front and own the item.

For businesses, the calculation is different. Maintaining appliances, or cars, or coffee machines, means redirecting resources (that you're already paying for) away from building widgets to something that isn't your core competency. Plus you get a bit of a discount because you're working with pre-tax capital.

Again, it's why car leases make sense for business, but are a terrible deal for consumers.

Paying $10-20/mo for a fridge is not going to significantly impact retirement savings over the long term. The alternative is putting that little bit extra into savings every month to buy the next fridge when it breaks. It balances out.

But your fridge is not the only thing you own. Presumably you would do this for all your appliances.

Keep in mind, that $20 a month for 40 years at 4% interest is $23000. If you add in more appliances this scheme can add up to be surprisingly expensive.

>Paying $10-20/mo for a fridge is not going to significantly impact retirement savings over the long term.

First, your numbers are way too low. Nobody is going to lease you, for example, a $2500 fridge where the capital cost will take 20 years to be paid back, especially since cost of service, depreciation of asset, and profit, needs to be priced in as well. You'd be looking at something on the order of $50/mo.

The opportunity cost of putting $50/mo into a tax sheltered investment account (with a reasonable 6% return) is around $15,000 after 15 years. In other words, you paid $15k for a fridge that retails for $2500.

> The alternative is putting that little bit extra into savings every month to buy the next fridge when it breaks. It balances out.

Huh? If you're overpaying for an appliance, you don't really 'fix that' by saving more.

Exactly! It's what I see of the spacex model - if you standardize one model and do it well, you find the weak points and either reinforce them or improve them.

The problem is consumer electronics break way more often than business not because they are not designed for heavy duty, but because some consumers are likely to misuse / misplace / misconfiged it. Since the price of your monthly payment is ultimately determined by the percentage of failure. This will drive up the total price, and hence it wont be cheap.

I think it will work well for Rental Apartments with property makers where large quantities are purchased and with design / placement standardise. And those price are included in rent or management fees.

I just bought a dishwasher and the whole tub seems to be stainless steel. You just have to look at the higher end ones I guess. This one was a Bosch 500 series one.

Having some experience in hardware engineering, I can tell you that I'd never touch consumer hardware. Far more profitable to build for industry, B2B sort of stuff. Sell pickaxes.

I think the hardest part of hardware is that its development is essentially an optimization problem - find the cheapest way a factory can make a widget with the features you're trying to sell.

The hard part of this problem is that there's an immense amount of information asymmetry between factories, designers, and marketers.

This will get better with factory automation. Right now you can upload CAD files to get a quote. The issue for designers is you want the quote to tell you why something is expensive (eg, a feature requires three tools for machining a part instead of one, or it's slightly too large/small for a process that is way cheaper, etc). Normally you work through this with engineers, designers, and factory staff iterating together on designs before tooling. That process is expensive and time consuming - it's a very human problem that can and will be automated soon.

I don't see "hardware as a service" coming into vogue for markets that are a race to the bottom (except if the service is sold to advertisers, like with Rokus). I see the service being manufacturing the hardware itself. It's already turning that direction.

I've seen plenty of companies get screwed by thinking they need to optimise BOM cost above all else. They make decisions that save a couple of dollars / unit but blow out their NRE costs and shipping dates by months or even years.

HaaS was my goal with http://www.gameref.io (selling/renting to eSports tournament organizers) -- unfortunately, and as others say in this thread, I couldn't raise enough money to move out of the prototype/MVP stage. Hardware is hard, and even with working my butt off to get covered by PC Gamer, Tom's Hardware, Vice, and dozens of others, I couldn't get enough nibbles.

C'est la vie.

= renting?

People owning stuff is less profitable. See: software, media, hosting.

Reply on news.ycombinator.com