Ask HN: What are the signs a piece of technology will become commoditized?
And what are some current examples? I know it isnt a "technology" in the traditional sense of the word, but i had a front-row view of Interest Rate Curves over the past 15 years. They went from arcane numbers some PhDs calculated at the back of the trading floor, to things Business Analysts started calculating, to things you could calculate with a standard website to things that were completely transparent and commoditized. In 2007 when I started building an interest rate curve system for a large asset manager, it cost us about $1.5m to get it up and running and another $200k/yr to maintain/operate. A lot of it was about getting clean data, but there was a lot of model tweaking/testing/maintenance as well. In 2014 when I left the same Asset Manager, we were able to re-do the same project from scratch with a $100k vendor product. Many quants got rich carefully guarding the creation of "The Curve" and were upset when it became easy to calculate w/o expensive workers. What lead to the commoditization? Several things:
1 - Academic review - lots of papers came out about how to do this. it wasnt patented, it was just a trade secret, but no longer...
2 - Open source, especially QuantLib, which implemented what the papers described
3 - Reference implementations - people outside the industry trying to break in released many reference implementations
4 - Overseas workers - people outside Wall St wanting to break in started doing it for free in hopes of proving their skills
5 - Democratic forums - people on the outside shared knowledge freely, helping unravel the internal trade secrets. In its early stages, every technology is vertically integrated, aka tightly coupled. If you were to buy a minicomputer or mainframe in the 1970s, or a personal computer in the early-1980s, the same provider would make the hardware, peripherals, OS, applications, etc. As the technology develops, that changes -- the PC market in the 1990s turned the hardware, peripherals into commodities. Until Apple's resurgence, it looked like the whole market outside of the OS would be commoditized. With smartphoness in the late-2000s, it was a very similar story: tightly-integrated early phones, gradually maturing into more commoditized hardware, peripherals, and applications, while retaining a premium market segment (iOS) with tight vertical integration. This pattern is fairly similar across industries, some good reading on the topic: https://hbr.org/2009/12/vertical-integration-can-work/ Clayton Christiansen, "The Innovator's Solution", Chapter 6: "How to Avoid Commoditization" Simon Wardley and "Wardley maps"; a good intro is here: http://www.slideshare.net/swardley/general-2013 I think HP was seriously hurt by this, losing to Intel in R&D right? was The Innovator's Solution a good book ? That chapter, and the two immediately preceding and following, were excellent. The rest was all right. I'd think that the biggest sign (though arguably a sign that a technology has already become commoditized) is when you see a leading vendor explicitly announcing/advertising a price cut. At that point, the story they're telling is concretely shifting away from the narrative of "our technology is uniquely suited to your needs" and toward "we can give you more widgets per dollar than the other guy". Example: GoPro. They recently announced a substantial price cut, and their stock took a pretty big hit in the wake of that announcement. Absolutely. And once the leading vendor cuts the price, it becomes a race to the bottom. I don't mean this quite in the 'derogatory' way it's going to sound like, but offshoring is a serious sign of trouble when it comes to commoditization. If what you're doing as a business can be done for a fifth of the cost in the Philippines, India, etc, then you have a serious clock on your business model before the 'quality, quality, quality!!' argument runs out. Seen it with a few things now. The biggest flaw that businesses do in this situation is believe in that the quality of the product will continue to outweigh the cost difference. It won't. ;/ By the way, I'm not saying there's inherently something wrong with the quality of work from those nations, I'm speaking strictly about offshore 'hubs' that are set up to purely be a cost-saving. Those type of hubs have a hugely different mentality when it comes to production in my experience. "The biggest flaw that businesses do in this situation is believe in that the quality of the product will continue to outweigh the cost difference. It won't. ;/" Do you count the loss in sales as cost? Or time to rework something? Sometimes it's hard to figure out if there's going to be a correlation before you do it. About 10 years ago, we tried to do most of our work offshore. Company said they could hire 5 engineers for the price of one here in the states. We started looking overpaid and I'm sure the managers thought we were. It didn't last long. We still offshore, but it's much more strategic. Sounds like a very familiar story to ours. Point being, commoditization is not a binary. It's not on or off, there's varying degrees of just how bad it is. At some point during that scale, the quality argument tends to fall down a little bit and I've seen it happen a few times. Interesting to hear you still offshore though, despite obvious troubles. We had offshoring before. But the push 10 years ago was to do the bulk of work offshore. Like to have the lead here working with 5 developers there. I can still see it working though. 5 to 1 is hard to ignore (although I think it's like 3 to 1 now). It comes down to implementation. Like everything else. Every technology will become commoditized sooner or later eventually. But here are some signs to tell when a technology goes from an innovation to a commodity 1) An increase in competitors is a sure sign of commoditization. Increase in competitors -> price war mostly -> existing core technology becomes commoditized as players start to differentiate with value-added services/products. 2) A new technology emerges that's 10x better on some dimension but still out of reach and a few years out of most customers' hands. It forces existing market for the technology to demand lower and lower prices for old tech -> commoditization of that tech as original innovator no longer might find it worth it to pursue that product. 3) Regulation could some time force commoditization. Often seen in healthcare/drugs production. I'm sure there are others. The common vain is that there's a rapid increase in competition. - When there is a bunch of people on the internet discussing it. - When you can buy the components on Alibaba for a fraction of the publicy known price Example: electric powertrain for cars. You could convert your car in a pure electric one for less than 4000USD by buying the motor & controller on the internet. And 22kWh worth of batteries (18650) for less than 6000USD. Very interesting stuff! >for less than 4000USD I'm assuming this isn't including paid labor and you're doing the engine/tranny swap yourself? Do you have any links to people doing this? Doesn't all technology eventually become commoditized? At the latest, after the patents expire. I think every piece of tech ever invented is either commoditized or is on a path to become so. I think it would be more productive to see if anyone can name a tech that isn't commoditized and isn't likely to in the future. There's some discussion of companies that hold monopolies in free markets here: http://www.fool.com/investing/general/2012/04/18/for-monopol.... Basically, small companies with strong brands in slow-growing, niche industries. There's nothing really assuring their monopoly except that it's not worth trying to compete with them. Regarding tech, the best example I can think of is TI and graphing calculators. They don't have a monopoly but they have a strong grasp on a niche industry - it's far from commoditized. I'm sure a company could create a clone, but I don't know if it would be worth the effort for such a small amount of revenue. Nuclear power. Most things military. Many things aerospace. You could make an argument that those things aren't bought and sold in a free market. Passenger planes and general aviation planes are at least semi-commoditized, but fighters, bombers, spy planes, missiles are all built for basically one buyer, the US DoD. The DoD decides which firms succeed and which fail, essentially. The "market" for nuclear power plants is so regulated as to prevent it from being "free" except to someone whose job depends on that market being "free". Considering the huge risk someone would be putting many others in by attempting to construct and run a fission facility, I'm not sure anyone should be "free" to do so. What's your definition of "free?" I'm possibly free to construct and run a fission facility, but I still have to follow the relevant regulations at all steps of the process. But a coal plant spewing radioactive by-products into the air is perfectly kosher, amirite? No, I believe you are wrong. Good, because it sure sounds wrong to me. Otherwise, I'd have to believe that radioactive pollution is only a problem when it looks bad on CNN. I wouldn't call the market for general aviation semi commoditized. Even the state of is working hard to create a decent competitor. Dead media is not (or rather, is no longer) commoditized. (By dead media I don't mean something that is vaguely less popular than something newer but is still produced, like newspapers, but instead something that has been superceded entirely, like Qipu or the radio-iconoscope.) It's easier to answer in reverse: businesses escape fierce competition (the normal state in a competitive marketplace) if they have - 1)Trade secrets, such as aluminum casting (if I recall correctly) for Musk and SpaceX, and Intel's tempering of memory chips in the early days. 2)Patents - when others can reverse engineer what you've done, you don't have the option of 1). Copyrights, somewhat similar and can be a part of tech products in the US (code, etc.) 3)Network effect/high cost of entry. (Most likely when monopoly and competition laws are poorly enforced, as is often true in new fields.) But disruption is a possibility. 4)Law/corruption. More common than thought. The EU thinks the US is simply ignoring antitrust laws re its large companies such as Google (anything free can count as bundling), for example. Interoperability isn't even on Facebook's radar. This can take a fairly soft form EU laws forbid wines from outside a given region being called by the famous name (such as "Champagne") and may soon forbid overseas data storage or passage too. Security laws in China, re tech products, will likely restrict foreign competition there in the long run. Amazon is being allowed to in effect enforce ebook prices of at least $3 with no action from the justice department, leveraging that policy to gain exclusive access to large numbers of books (if they want any free days.) Strictly illegal, but it's just not in the US interest to go after their own Hegemonic corporations. So ebook distribution, the most obvious of commodity services... turns out not to be a commodity at all, somehow given Justice department inaction. There are other temporary causes, such as a shortage of workers with particular knowledge or skills (coders just now.) Well, here's hoping that's temporary. Refusing to participate in a race to the bottom (in quality say, and taking a hit for a while) can also leave a company without obvious competition for a time. I've left out cornered markets on minerals, not too likely at the moment. One could also add patent pooling, which US automotive companies used both to save on research and raise the cost of entry, in the short run in order to create an oligopoly. Peter Thiel also has a good description of how to differentiate (avoid competition), through either:
1. Proprietary technology
2. Network effects
3. Economies of scale
4. Branding
You can read the relevant chapters from the book 'Zero to One' on genius.com: http://genius.com/Chapter-3-5-of-zero-to-one-peter-thiel-ann... Many products/techologies follow the S-curve. Slow start, exponential growth, plateau (mature product, 3-5% growth per year) - and then slow death. From my perspective, we have a broad set of capabilities. Let's take transportation. There are performance attributes like time, speed, work, power, mechanical efficiency, etc. The wheel was a technology that was surpassed by the horse and buggy and then eventually the automobile. The capability is needed but the technology or family of products eventually becomes obsolete. Commodity products are on the plateau. The technology has been proven and there is broad market acceptance. The unit cost has been brought down substantially from the point of market entry. Tesla's electric vehicles are pushing in this direction but it will still take some time. When open source implementations, reference guides/manuals and public internet forums (in that order from biggest to smallest sign) start being ubiquitous. Mobile Device Management (MDM). Blackberry got the ball rolling in the early 2000's, but it was the release of iPhone and Android that got vendors rich. Apple and Google created API's and enrollment processes to manage devices (device lock, wipe, etc.). Companies created software to consume those API's, the biggest being AirWatch, Mobile Iron, and MaaS360. AirWatch was acquired by VMware for $1.5billion in 2014 and Mobile Iron went public, rising 22% in the first day of trading. MaaS360 was acquired by IBM for several hundred million. At first, license fees were high, competition wasn't too fierce, and the market was booming. Then many more low-cost players moved in. Open source MDM software appeared. One company (Miradore) gives MDM away for free to hook customers on IT management fees and various other upsells. Mobile Iron stock is down 70% for the year. I was listening to an NPR podcast this morning and heard an advertisement for a product named Bushel doing Apple MDM. I think the definition of commoditized tech is when you hear companies advertising on NPR. My take on the situation is that it's risky to build a core business merely consuming other company's API's. Doing analysis and more advanced things, sure. But simply executing an API command is not special. The MDM list of features is completely controlled by Apple and Android manufacturers, and anytime they release a new feature, every MDM player instantly gets access to it. The only differentiation between products becomes how effectively they use the pre-existing API's (scaling, integration with other enterprise products, etc.) and pricing. This obviously becomes a race-to-the-bottom on pricing, destroying margins. The reason I know so much about this is that I work at a company called Apperian, focusing on Mobile Application Management (MAM). We use to compete with the MDM guys, but it's a completely different game now. All of our core technology is unique, most of it is patented / patent pending, and we solve problems that the MDM API's cannot. We dominate the MAM industry, although niche and much smaller than the MDM sector, is growing very quickly. It would be extremely difficult and costly for a company to implement what we do, in contrast to the MDM guys. It's been very interesting to watch the rise and fall of big players, as our sales keep booming. I will keep watching the drama unfold. I had to Google MAM to clarify what it meant, and then I realized this was what I was wondering about the other day, puzzling over how BYOD isn't a massive security nightmare. I must admit, I'm very fascinated with application management would actually be executed, from an academic / computer science angle. I can also completely understand patenting from a business perspective, but I'm also curious as to what things have actually been (publicly) patented. How is MAM different from MDM? MDM uses operating system level API's provided by the vendors (Apple, Google). These require the device to be enrolled with management software (Airwatch, Mobile Iron, etc.). MAM vendors like Apperian use various other techniques to manage the applications and their data only. The device is completely irrelevant, because we only care about the app. We do this through clever use of API's and a technology called 'App Wrapping'. This takes the mobile app, decompiles it into machine code, and very carefully injects new security features directly into the decompiled app. This allows adding security features like data at rest, analytics, etc. without needing any OS level API's. https://www.apperian.com/mam-blog/app-wrapping-lets-get-tech... Wow, thanks for the reply! :) > decompiles it into machine code ...this was the part my eyes started getting big... > very carefully injects new security features directly into the decompiled app ...and this was the part I had a small jawdrop moment. :P > This allows adding security features like data at rest, analytics, etc. _Wow_. I think I suddenly get the rationale behind patenting. Hah. * Reads link * > Mobile app wrapping is able to provide the same level of security as device management and even takes it a level deeper by securing the app itself and inspecting that app to know what it is doing and if it contains any malware. (Italics at the end mine) After an admittedly short look around, I've concluded that you seem to be focusing entirely on enterprise-controlled system-/app lockdown. To me, the technical accomplishments you've achieved to make this possible provides a massive opportunity just waiting to be fulfilled: a new, unique approach to mobile antivirus. It wouldn't be for everyone, but I can totally say that I'd use it, and a lot of technically-oriented "tinkerers" would probably love some of the introspection and analytics your system provides, in an antivirus context. :D I suspect most of your existing tech could translate across really easily - and that the effort you make to make antivirus better will also translate across into even better lockdown heuristics. Granted, I don't fully understand the platform, so I might've missed a point somewhere, but I still think this is pretty amazing tech. Thanks for the info! :D Are you looking for a job? We are hiring now for solid engineers. seibeljames@gmail.com or jseibel@apperian.com. We are doing some very interesting tech for governments / corporations with high-security clearance needs. I can explain much more in depth if this interests you Volume leads to low cost and high quality. That's why Ethernet got commoditized but fiber channel didn't. Today notable examples are in solar power and batteries, but also think sparc vs x86 vs arm (mobile volumes much higher than desktop/server) On the supply side:
No patents, or easy/cheap to license patents. Technical trade secrets don't remain so for very long.
Large margins for one or a few suppliers/manufacturers. This is basically what is happening with the smartphone business. First Apple started it. Then there were a few players like LG, Samsung, HTC, etc. Now everything and their uncle is starting a smartphone company. Companies like Lenovo, video con, not to mention the Asian upstarts from China and India which are definitely late entrants to the market. On the demand side:
Steadily increasing users:
Early adopters not abandoning the product, but continuing to use it.
Increasing numbers of early adopters in various social circles. I guess opensourcing is the sign. See what Joel has to say on this: http://www.joelonsoftware.com/articles/StrategyLetterV.html "Commoditized" meaning "sold as a product?" If you can produce more than one and sell it at a price people are willing to pay, it will be commoditized - right? Particle colliders probably won't be commoditized; spacecraft will. It usually means a product which can be bought from multiple providers, all largely equivalent in terms of features, and therefore primarily competing on price: https://en.wikipedia.org/wiki/Commoditization Crude oil and oil paintings signed by Pablo Picasso have both been known to be sold for money, but only one of them is a commodity. If a technology is useful, I think it's guaranteed that it will eventually become commoditized. But as to the process, check out Simon Wardley's work, including Wardley Maps: http://blog.gardeviance.org/2015/02/an-introduction-to-wardl... When you see people telling in public about talent sortage in the XXXXX field/sector/technology - you can bet this is about to be commoditized. Why? because the ones that send that message need/want as much cheap labor on that field as possible, and that's the way they can get a lot of people moving on that field. If the new technology can help an emerging industry/market to make things cheaper, easier, or with higher performance then the technology will become commoditized to be adopted in large scale. The time it takes to be commodotized depends on competition, and stage of industry life cycle. The only reason everything doesn't end up commoditised is because they're protected in some way. Like, Facebook having all the people, or patents. If not, it's a race to the bottom every time. You should reverse the question and look for barriers to entry. I'm no marketeer, but off the top of my head: - Branding
- Word of mouth
- Market value It's tough to become commoditized in this market tho' It seems like people are assuming that all products are commoditisable from a market point of view, and from an industry point of view, which isn't true. E.g. Just because watches are commoditisable from an industry point of view, in the sense that they can be manufactured en-masse for next to nothing, the buying behaviour by consumers for watches indicates that they can not be commoditised, because brand equity is a factor in purchasing, features of the watch are a factor for purchasing, etc, and therefore product differentiation stops them being a commodity. On the other side, just because consumers would accept commoditised, mass produced, cutting edge CPUs, does not mean that industry can provide this. The barriers to entry for this industry prevent them from being commoditised (R&D costs, human resources, capital costs and so on) So to look at what technologies may soon become commoditised, you need to analyse current and future demand, and current and future supply. You'd also need to analyse both the markets for the product, and industry groupings within the supply chain for that product. Some simple analyses might give you an idea, e.g. a proper Porters Forces analysis, consumer indices, technology trends, etc. An interesting current example for me are 3d printers. Because the barrier to entry for designing and creating and selling a rudimentary 3d printer is so small, it has the potential to be commoditised. But from a marketing point of view, it probably can't, because it's a new technology and so brand differentiation might affect buying behaviour. From a consumer perspective, not enough people want or care about 3d printers, so in turn, suppliers wont churn them out. From an industrial point of view, it would be hard to create barriers to entry for new entrants to the market because it's not complicated to create a rudimentary one, so it's probably not a profitable venture for incumbents, and so on. Another example would be the resurgence of VR. That's an interesting one because you can witness companies doing their best to create barriers to entry for new competitors, to avoid it being a more commoditised product, despite also being easy to create a rudimentary version (emphasis on 'rudimentary'!) From a marketing fundamentals point of view though, commoditised products tend to be ones that are cheap to make, are consumed in some manner, and which consumers don't have a brand preference for, therefore buying the cheapest product available. E.g. screws, staples, but not, say, certain breakfast cereals (because people exhibit preferences for different brands of cereal and don't just buy the cheapest possible). Most new technologies can't fall within these parameters. When people start mass advertising to do it outsourced. payment processors To some extent. You can write a nice layer over the top but fundamentally underneath payment processors are controlled via Visa and Mastercard. That's one of the reason payment processors are so damn careful to descend like the wrath of an angry god when someone does/sells something Visa or Mastercard wouldn't like. Exactly. But now we have Bitcoin. There's non-zero probability that banks will switch to it under the hood[0] and there's especially high incentive for small ones that want to do international transfers[1]. Once we have common protocol, you can use whatever payment processor you want, and unlike now, you can create one in your garage. It's not hard to see that I'm biased towards Bitcoin, but even assuming it doesn't exist, sooner or later banking system as we know it today must change (dramatically). I would assume this will make companies like Visa and Mastercard much less relevant. [0] - http://www.coindesk.com/bank-of-america-cryptocurrency-wire-... [1] - https://www.reddit.com/r/Bitcoin/comments/3ncat3/why_bitcoin...