Makers Movement Deserves More Than Buzzword Hype
The era is defined by “IoT (Internet of Things).”
“Makers” is changing the world.
I agree with both statements. But when words start walking on their own, it’s fine to leave them alone — we just need to be careful not to get carelessly swept along.
The other day, for the first time in a while, I had a moment to sit and think idly.
It was really just a few minutes waiting for the bus at a bus stop. Lately my mind has been occupied constantly with work, my kids, side projects, and last month a two-week trip to Germany, plus the preparations before it. Going back further, I’d been far too busy to have a calm moment to think. Small pockets of free time exist, of course, but I usually spend those checking the news or catching up on the latest tech, some activity that’s half work and half not. Every time, without fail. So it had been a genuinely long time since free time appeared naturally, rather than being deliberately carved out.
And when that happens, this is what comes to mind:
“Since I have the chance, let me think about something that gets at the essence of things.”
So I turned my thoughts to a long-standing question: in what direction is the future of the hardware business heading?
As the keywords above suggest, the era is indeed changing, and steadily so. It’s probably fair to say that businesses built solely on internet solutions have saturated. Competition has intensified, profits have shrunk, and as a result the number of companies has kept dwindling. It’s not too much to say their valuations are trending downward too. This is an entirely natural flow of the times. That said, news still breaks about massive acquisitions that make you think “isn’t this amount insane?” But what I want to focus on here is the state of the market as a whole behind those flashy headlines, especially at the grassroots level. Meanwhile, hardware-related businesses are booming and drawing public attention. The background and examples fill plenty of books and online news already, so I won’t smugly repeat them here.
Let’s get right to the main discussion.
Let’s do “hardware startups” properly
The symbol of the Makers era is probably crowdfunding, epitomized by Kickstarter. Hardware development costs a fortune no matter what. Covering that with pre-order funding is a genuinely great idea, and on top of that it gave users a huge benefit: becoming an early adopter of an interesting product. In recent years, fewer companies had been taking on hardware challenges, but judging from the crowdfunding trend, it’s become clear that the market had a smoldering demand for hardware all along. But relying on crowdfunding, depending on how it’s used, can amount to nothing more than supporting a small business. And “startup” and “small business” are hugely different things — in terms familiar in Japan, the difference between a “venture company” and a “small/medium enterprise” is exactly it. A big difference, isn’t it? Small businesses can end up running on hand-to-mouth cash flow. Hanzawa Naoki isn’t just a TV drama, it’s a real story. Why does that happen? Because it’s hard to get upfront investment. Most small businesses launch on their own money — meaning debt — and afterward keep operating perpetually by stacking sales on top of that capital, paying back debt, taking on debt again. In that kind of scene, crowdfunding is truly a savior. Startups, on the other hand, must not be businesses aiming to earn $1 million or $2 million. They need a business model that generates billions, tens of billions in profit, and then keeps growing from there. Actually, that’s not quite the right way to put it: if you can’t envision that kind of big dream, you’re better off not doing a startup at all.
In other words, running a Kickstarter, launching a product, making decent money, and then stopping — that has no continuity at all. Naturally you’d use that capital to build the next product. But the development cycle for hardware products is long. Even in the fastest case the next release is six months out, and realistically one or two years out. A startup is a company that keeps growing. So what do you do during that one or two years?
Let me just write the answer here.
You use that one or two years to grow a service. Don’t misread the word. It’s not “launching” a service, it’s “growing” one.
Hardware is fundamentally low margin. With electronic appliances, for example, it’s normal for retailers to take roughly a 30%-50% margin. The product cost itself runs roughly 30% of the sale price. Shipping incurs transport costs, and crossing borders means it gets taxed there too. The tax rate depends on the product category and destination country, but it’s roughly 5%-15% of the transaction price. Internally, there’s personnel cost, and depending on the case, outsourcing costs pile up as well. So how much profit actually remains with the company that developed the product? There’s some case-by-case variation, but it’s mostly a single-digit percentage of the product price. Take a $95 product: to make tens of billions in profit, you’d need to sell roughly 100 million units. For a $470 product, you’d need to sell tens of millions of units — iPhone- or iPad-level volumes.
I want you to feel how unrealistic that is.
Meanwhile, “internet business is huge when it hits” sounds like a bubble-era phrase from decades ago, but it remains a plain fact even now. The reason is low initial investment combined with a kind of favorable uncertainty. Low initial investment goes without saying, but ongoing investment is naturally low too. (I know of a startup that ran on just two PCs and $600 a month in rent.) Dropbox recently launched what felt like a scorched-earth offer — $9.99 a month for 1TB of cloud storage, previously 100GB for the same price — and that too reflects a different sense of cost. For them, acquiring users is the supreme mandate, and few people inside the company are sitting there calculating the per-byte overhead against the break-even point. A few probably are, but that’s clearly not their main concern. And in many acquisitions, the price seems to be decided with almost no relation to revenue. A commonly cited formula is “(number of users) × ($X) = (acquisition price).” Why does it come down to something so seemingly vague? Because when a service is growing rapidly, future value can’t be measured from current revenue or profit. So easy-to-grasp indicators like user count or download count become the yardstick. There’s surely more complex logic and negotiation behind the scenes, but that information isn’t disclosed, so there’s no way to know.
People who’ve worked in hardware, watching this glittering situation, might not grind their teeth, but they’d at least think the grass is greener on the other side. Honestly, in many respects, service businesses have had more freedom, more agility, more profit.
As a bit of an exception, one pattern where doing a Kickstarter and then stopping can still grow the business is patents. That said, this is an area where it’s hard to draw a clear line of causation to fundraising or product releases, so I’ll leave it out here. It would take us off track.
The era is tilting.
Whether you believe it is up to you. The pieces have already been given, and how you assemble them is up to each person, so the conclusion can differ too. But for those who believe it — only those who can believe it — please keep reading.
The era has come full circle
Buzzwords are things someone strategically coins with the media in mind, so being catchy, easy to understand, familiar, and easy to use matters more than capturing the essence. As a result, they often show only one facet of things. IoT is no exception.
Internet of Things — “the internetization of things” — is probably the plain translation. “Various things are going to get connected to the internet.”
Hmm. Isn’t this a phrase I’ve heard somewhere before?
Right. It’s what everyone was saying in unison around the time IPv6 was established.
But what actually followed was the era of smartphone dominance, and “everything gets integrated into the smartphone” became the new catchphrase. Digital cameras, video cameras, media players — all fell prey to that era. And soon, PCs too.
Why didn’t the IoT era arrive back at the time of IPv6? That’s a story from roughly ten years ago. There was no iPhone, no iPad, no iTunes Store, no App Store. There was no Dropbox, no Evernote, no Spotify. Facebook had only just gotten off the ground, and even Google hadn’t gone public yet. In other words, internet-based solutions were far too weak, so it was a state of “connect it to the internet — and do what with it?” No wonder it didn’t catch on.
But the era changed. Integration into smartphones settled down, and a kind of boredom with an unchanging platform helped give rise to a new hypothesis: demand for dedicated devices might emerge again. Some irresponsible people latch onto that and declare “now is the era of IoT.”
Well, of course some people do. I understand the situation. A few success stories have started to emerge too.
The most obvious recent example is probably Nest. Founded only a few years earlier, with just two main products, it was acquired by Google for roughly $2.8 billion — twice the price of YouTube. https://store.nest.com/ Nest’s annual revenue is roughly $280 million; using the earlier logic, its annual profit would be at most around $9 million. In other words, Nest’s revenue could never generate enough profit to pay off $2.8 billion.
And yet the acquisition happened.
Plenty of speculative posts cover the reasons and strategy behind it, so I’ll leave that research to you.
The extremely important point that tends to get overlooked here, and probably deserves very rigorous scrutiny, is whether the process of planning and developing hardware centered on an internet solution has been established, maintained, and properly operated. Nest achieved that beautifully. And as a result, it landed an acquisition price on a scale unimaginable from its revenue figures. This connects back to the earlier question of why internet businesses are huge when they hit.
Still, this line of reasoning stings quite a bit when it bounces back at me in various places. But pain is no excuse to avoid it. This really matters.
Platform
Let me raise another example: bttn. https://www.indiegogo.com/projects/bttn As an object, bttn is just a button, but by connecting seamlessly to the internet, it functions as a trigger for various services. You can call a taxi with bttn, or send a notification to someone with bttn — I once made a “husband-summoning Yo switch” for my wife, and she dismissed it flatly with “I’d never use that.” Lol. Personally, I think this is a product with real potential. You could say it perfectly matches the IoT keyword.
However, I myself don’t think bttn will achieve massive success. The reason is that it uses the IFTTT platform as the hub connecting the physical button to internet services. IFTTT is currently unstoppable, one of the services expected to grow rapidly, and bttn is riding on that momentum. https://ifttt.com/ What’s wrong with riding the momentum of a hot service? Let’s think about it. Suppose, in the future, bttn’s concept gets accepted by the market. Without a doubt, a flood of companies making similar things more cheaply will appear. How does bttn fight that battle then?
To put it a bit more bluntly: does a company that doesn’t hold the service platform even stand a chance?
Get dragged into price wars, worn down fighting off knockoff junk, and even if you somehow manage to secure a market share of a few tens of percent, revenue is so-so but profit is razor thin. Would that really be something to be happy about?
The keyword “platform” is extremely important in a company’s growth strategy, and it’s not a simple story. Consider a situation where Facebook acquires some company. Let’s think through a few examples of companies that might seem related to Facebook’s business.
Company A: a company making games on top of Facebook’s platform Company B: a company providing its own services or features using Facebook’s API Company C: a company running its own platform with the potential to become an alternative to Facebook
Company A would rarely draw any interest. Unless it’s making an unusually large amount of money, a company like A is just one of a bundle that appear and collapse, grow and shrink, bubble up and vanish. Company B deserves consideration, but an acquisition is unlikely. For a company holding the platform, B is a convenient partner testing out “the future” on its behalf. So the response tends to be one of two choices: form a partnership if a meaningful relationship seems buildable, or copy the idea and crush the company if the service looks promising. Apple does this constantly. Apple built the iOS platform, and by offering the App Store as a playground, succeeded in gathering an enormous number of samples. Amateurs and professionals alike built countless apps, and Apple was in a position to observe the entire trajectory of their sales. In other words, the future the platform should take became fully visible to Apple. Then Apple picks out a few promising apps and absorbs their features as Apple’s next default feature. The original developer gets crushed. This is outright copying, nothing but exploitation of technology and ideas, but for a company providing the platform, it’s justified as a perfectly ordinary act. A frightening thing. Company C is the important case. It absolutely must be crushed or absorbed, and in most cases crushing it is extremely difficult. If it holds some advanced edge and stands in a superior position, catching up and overtaking it is far from easy. Good examples are WhatsApp and Instagram. While Facebook’s active users trended downward, these companies growing their user base were nothing but a threat to Facebook. The one area where Facebook was clearly superior compared to these vigorous up-and-comers was money. So, acquisition. There was probably no room for choice.
Looking at it this way, only Company C has a glittering future waiting for it. Of course, turning down an acquisition offer and aiming to take over the world outright would also be its own kind of fun. The key difference between Companies A and B versus Company C is whether or not they built their own platform. That changes the growth story into something entirely different.
Naturally, this is a pipe dream. Holding a platform is not something achieved easily. Most companies will fall partway through the dream.
So do you give up? Do you settle for so-so?
Let me return to the earlier words.
“Startups must not be businesses aiming to earn $1 million or $2 million. They need a business model that generates billions, tens of billions in profit, and then keeps growing from there. If you can’t envision that kind of big dream, you’re better off not doing a startup at all.”
Let me talk about “money” in a way that makes clear what a startup actually is. Take what follows purely as one example, one facet. A term often used in startups is “burn rate” — at what pace do you burn through capital? Concretely, it’s usually used to mean how much money is being spent each month. In a typical startup, at the seed stage it might be modest, but once you’ve gone through Series A and Series B funding, even with zero revenue, a burn rate of tens of millions of yen a month is entirely plausible. Nothing to be surprised about. Why does it get like that? The answer is simple: personnel and equipment are all spent at “the best,” and the aim is “the best” in both speed and quality. Generating profits in the billions or tens of billions is no small feat. Clear differentiation from other companies, competitive advantage, patents — all sorts of things need to be built up in a short period. That takes a great deal of money and effort, and while compromise is of course possible, doing so means it’s no longer a startup. That’s what most people involved in startups believe. Can a company launched on meager personal savings pull that off? Can it borrow hundreds of millions of yen from a bank as a year’s worth of operating funds? There’s a clear line drawn there, but in any case, this is only possible with the support of social infrastructure.
I’ve thought deeply, in a separate post, about what infrastructure — the ecosystem — should look like for startup development in Japan, so please refer to that.
The mismatch between the two buzzwords, “Makers” and “IoT”
Another keyword tied to Makers is probably 3D printing. The technology behind 3D printers has been around for a long time, but it finally became affordable enough for ordinary people to reach. That was a major revolution: it made it possible for individuals to prototype physical objects freely and easily.
The recently popular flow for launching a hardware business goes like this.
- A business/product idea comes to mind
- Experiment on a breadboard for now
- Draw a schematic and build a PCB. The enclosure can just be 3D printed
- Shoot a video
- Raise funds on Kickstarter
- Use that money to build molds and go into production
People say, “Wow, it’s gotten so much easier to launch a hardware business. A new era has arrived!”
It’s true that steps 1 through 4 have become dramatically faster. Beyond 3D printers, various technological elements — tools that streamline prototyping like Arduino, the falling cost of modules, the mainstreaming of mobile battery technology — have all come together wonderfully, as if converging on exactly this era. For a project with no technically difficult challenges, these steps are probably achievable in a month or two, much like launching an alpha version of a software service. (A famous story is that Tumblr apparently went from concept to first launch in two weeks. There are plenty of similar cases.)
But please think carefully about steps 5 and 6.
For someone consciously avoiding the hardware business, the biggest concern is probably the high upfront investment. I once saw an article claiming that Apple, while making the iPhone 4, spent $1.7 billion building a dedicated facility to measure radio characteristics. That’s an extreme example, admittedly, but measurement equipment, prototypes and their disposal, mold investment, personnel costs — hardware involves a huge amount of spending before it ever reaches the customer. Even for a fairly low-cost, technically simple product, it rarely comes in under $9.4 million. But raising that much for such a boring product wouldn’t make sense, so for a product with some novelty, some advantage, some technical challenge, and a reasonably higher price, you’d still need at least $2.8-3.8 million in funding before the project could run smoothly. Keeping that in mind, look at the amounts raised in Kickstarter projects: projects raising over $3 million can be counted on one hand.
https://www.kickstarter.com/discover/advanced?category_id=0&woe_id=0&sort=most_funded
In other words, for most projects, the money raised through crowdfunding is really just an add-on to existing funds, or crowdfunding itself is part of market research, or it’s merely a trigger for pulling in new investment or acquisition talks. In most cases, whether the cash raised through crowdfunding exists or not doesn’t fundamentally change the story.
This phenomenon is a plain fact, and there’s nothing wrong with it. Crowdfunding’s meaning hasn’t been lost. Crowdfunding, to begin with, is built on the motivation of supporting a company that hasn’t yet reached product release, in exchange for getting an early product. So it’s purely a question of what the output of that support looks like — it doesn’t change the fact that the company is doing its utmost to live up to that support.
What matters here is that this chain of reasoning:
“Hardware business has gotten easier” → “If you’ve got an idea, launch it right away. There’s Kickstarter” → “The era of hardware startups has arrived” → “IoT is here!!!”
does not follow. That reasoning is wrong. The mistake sits right at step three: these are not hardware startups, but in many cases, hardware small businesses.
The practical stage
In discussing IoT’s definition, I touched on the matter of “service,” and pointed out that the definition of hardware business is shifting. Given that, standing as a “startup” takes more than scraping by on razor-thin funds, shipping a product, and then rushing out variant products to secure ongoing revenue. That won’t do. Release a service alongside the product, and by the time the product has gone through one cycle, have the service foundation solidified. Once you’ve secured enough users, invest the revenue into developing a new product. Isn’t that the kind of cycle it should become?
What’s needed for that isn’t whipping up hardware quickly in a short period, though that helps in some sense. What should matter more is taking time to thoroughly deliberate on the coordination between hardware and service, on the value born from that coordination — on the essence hidden behind the word “IoT” — and building several prototypes and testing hypotheses along the way. Picture a more concrete situation: suppose there’s a service idea but it hasn’t been launched in time, and only the hardware exists. Should you run a Kickstarter? No — that’s exactly when you should deliberately delay development.
As mentioned earlier, the barrier to hardware development has come down. Because of that, if you try to compete on hardware alone, a rival can easily release a copycat product and crush you. The real value you want to convey to users never gets across, and a wonderful plan ends up buried among a pile of worthless knockoffs.
The same holds for services: repeated test runs surface problems. And when the product is tied to an internet service, those discovered problems may need to be fed back into hardware development. The old assumption that the service can wait until the hardware is done no longer matches the product development process the coming era demands. More concretely: even without hardware in hand, you should work out the service specifications, and by the time the hardware prototype is ready, the service prototype should be ready too, so you can run local operational tests. Feed the discovered problems back into the hardware, and also extract requests toward the service from the hardware side. That’s the kind of flow it becomes.
This is an extremely important shift in process, and many people seem to sense it without fully grasping it. They say “I get it” but don’t put it into practice. Yet it’s clearly the watershed. I’ve cited Nest and bttn here, but there’s no end to examples if you go looking.
Of course, everyone understands the flow of the times. Some of you, even reading this far, might feel irritated thinking “you don’t need to tell me that.” (Then again, if you were that irritated, you probably wouldn’t have read this far down. Lol.) I understand that people get it. But how to put it into practice, how to ride it, how to exercise self-restraint — that’s extremely difficult. When drafting a business plan, you probably think up all sorts of service developments too. It’s fine even if it’s mixed with fantasy. Let your imagination fly, and you might even feel the whole world being repainted. But if you hold onto only that fantasy, and once the hardware is actually built you think “alright, let’s turn that fantasy into reality,” it’s too late. You won’t make it in time.
“We made the hardware for now. As for the service, well, we think we can come up with various ideas” is guaranteed to die young.
So what do you do? As a rough principle of action:
If there’s no clear service plan, it’s better to hold off on hardware development. If the service prototype hasn’t reached a satisfying quality, it’s better to postpone finalizing the hardware.
To roughly that degree, product development going forward needs a completely reoriented mindset.
As always, I’m not trying to offer anything or change the world here. I believe things carry more weight than pens or words. That’s why I do hardware work.
So this is really just a note to myself, a signpost I’m leaving behind for a time when I’ve forgotten it, or a time when I’m flustered.
If it happens to strike a chord with someone out there, I’d be honored. And if the opportunity arises, let’s work together. Lol.
Originally published in Japanese at https://clazytech.com/2014/09/91/. Translated with LLM assistance and reviewed before publication.