close
tech

Complementary and V-type strategies: Why are commercial companies so willing to invest in open source software?

ct9hvahbq8o7wa8l.jpg

神译局是36氪旗下编译团队,关注科技、商业、职场、生活等领域,重点介绍国外的新技术、新观点、新风向。编者按:微观经济学的一个简单概念可以解释科技圈已经发生和正在发生的很多根本性转变。这个概念就是互补品。所谓互补品是指客户一般会同时购买的两种产品。科技公司很多都会针对互补品展开此消彼长的V型战略:也就是对互补品进行商品化。被商品化的互补品价格下降后就会推动主打产品的需求上涨。比如IBM做微机的时候对插件商品化从而带动了对PC需求的增长,然后微软把PC市场商品化从而带动了对操作系统的需求增长。Joel Spolsky通过各种案例分析了这种产品战略。原文标题是:Strategy Letter V读大学的时候我选了两门经济入门课:宏观经济学和微观经济学。 宏观经济理论充斥着类似“低失业率导致通货膨胀”这样的理论,那种理论从来都没有完全符合过现实。 但微观经济学的东西又酷又实用。里面有很多关于供需关系的有趣概念,而且确实有用。 比方说,如果你的竞争对手降价,除非你也跟进,否则的话对你的产品的需求将会下降。本文中,  我会展示其中一个概念,这个概念可以解释很多我们所熟悉的某些计算机公司的事情。 在这过程中,我注意到开源软件的一些有趣之处,比如说:大多数公司在开发开源软件的事情上投入重金其实是因为这对他们来说是一项很好的商业策略,而不是因为他们突然间不再相信资本主义,反而爱上了 自由言论 。市场上的每一件产品都有替代品和互补产品 。 替代品是你可能会买的另一种产品,但前提是第一种产品太贵。 鸡肉是牛肉的替代品。 如果你是一个养鸡农民,在牛肉价格上涨的情况下,大家会想多买点鸡肉,这样你就会卖得更多。互补品是你通常会跟另一种产品一起购买的产品。 汽油和汽车是补充的。 计算机硬件是计算机操作系统的典型互补品。 幼儿照顾者是在高档餐厅吃晚餐的互补品。 在小镇上,一旦当地的五星级餐厅提供买二送一的情人节特别优惠,当地的婴儿照顾者就会把价格翻番。 在其他条件相同的情况下,当互补品的价格下跌时,对某产品需求就会增加。我再说一遍,因为你可能没注意,这一点很重要。 当某产品的互补品价格下跌时,对该产品的需求会增加。 例如,如果飞往迈阿密的航班变便宜了,那么迈阿密酒店客房的需求就会增加——因为有更多的人飞往迈阿密并且需要一个房间。 当计算机变得更便宜时,会有更多的人购买,并且犹豫计算机都需要操作系统,因此对操作系统的需求增加,这意味着操作系统的价格可能会上涨。说到这里,大家相当常见的一个说法是通过说“啊哈! 可是Linux是免费的啊!”,这是一种常见的混淆视听的说法,好吗? 首先,当经济学家考虑价格时,他们会考虑总价格,总价格包括一些无形的东西,比如安装所需要的时间,对每个人的重新教育, 以及对现有流程的转换。 所有这些我们喜欢称之为“总拥有成本(TCO)”。其次,通过利用免费啤酒作为论证,这些倡导者设法相信自己不受经济规则的约束,因为他们有一个很好的零,可以乘上一切。 这里有个例子。 当Slashdot   问   Linux开发人员Moshe Bar未来的Linux内核会不会与现有的设备驱动程序兼容时,他回答说不需要。   “专有软件经过调试的代码每行的费用为50-200美元。开源软件不会有这样的价格。Moshe continues to claim that every Linux kernel version can make all existing drivers obsolete, because the cost of rewriting all existing drivers is zero. This is completely wrong. He basically said that spending a small amount The programming time makes the kernel backward compatible, which is equivalent to spending a lot of programming time rewriting each driver, because the two numbers are multiplied by their “cost”, and this cost he believes is zero. It’s a prima facie error. The cost of hundreds of thousands of hours that developers have to modify each existing device driver must be at the expense of what. Before that, Linux would once again become disabled in the market because it does not support There is hardware. Is it better to use all those “zero cost” efforts to make Gnome better? Or to support new hardware? Debugging code is not free, whether it is proprietary or open source. Even if you don’t pay for it, it has the opportunity cost and the cost of time. Open source has a limited number of volunteer programming talents to help, and each Open source projects compete with other open source projects for the same limited programming resources, and only the sexiest projects really have more volunteer developers than they need. In short, I am trying to prove the economics of free software. People don’t have a deep impression, because in my opinion, they have made a mistake of zero. Open source can’t escape the economic law of gravity. We can see from Eazel, ArsDigita, and many other attempts previously called VA Linux. But there are still some things that the open source community can really understand: many very large listed companies, despite their responsibility to maximize shareholder value, still invest a lot of money to support open source software, usually in the form of large The programmer team pays for the latter to do open source software. The complementary principle can explain this phenomenon. Again: when the price of a product’s complementary product drops, the demand for that product increases. In general, the company’s strategy The benefit is to make the price of your complementary product as low as possible. In theory, the lowest price that can be sustained is at most Commodity prices—that is, the prices that appear when you have a bunch of competitors that offer goods that are not clearly differentiated. So: Smart companies will try to commercialize the complement of their products. If you can do this The demand for your product will increase, and then you can earn more money. When IBM designed the PC architecture, they used off-the-shelf components instead of custom components, and also recorded the interfaces between the components in detail. Into the (revolutionary) IBM-PC Technical Reference Manual. Why do they do this? This way other manufacturers can join the party. As long as your interface matches, you can use it in your PC. The goal is to commercialize the plug-in market, which is complementary to the PC market, and they have been very successful. In a short period of time, companies that offer memory sticks, hard drives, graphics cards, printers, etc. have sprung up. And cheap plug-ins mean that the demand for PCs will increase. When IBM got the PC-DOS operating system license from Microsoft, Microsoft left a heart and did not sell exclusive licenses. In this way, Microsoft can offer the same license to hundreds of other OEMs such as Compaq that use IBM’s own documentation for legitimate IBM PCs. Microsoft’s goal is to commercialize the PC market. Soon the PC itself has basically become a commodity, the price has been declining, the ability has been continuously improved, and fierce competition has made it extremely difficult to make a profit. Of course, low prices will increase demand. The increased demand for PCs means an increase in demand for its complementary MS-DOS. The other things being equal, the more demanding a product is, the more money you can make. This is why Bill Gates can buy Sweden and you can’t. This year Microsoft is trying to repeat itself: their new console Xbox is starting to use commercial PC hardware instead of custom parts. The theory is that the price of commercial hardware will drop every year, so the Xbox can also lower prices accordingly. 不幸的是,结果似乎适得其反:商品化PC硬件的价格似乎已经被挤压到商品化的地步,所以Xbox的制造价格并没有像微软期望的那样降得那么快。 微软Xbox战略的另一块是用DirectX,这个图形库可用于编写在各种视频芯片上运行的代码。 其目标是让视频芯片商品化,降低它的价格,好卖出更多的游戏,因为游戏才是真正赚钱的地方。可是 为什么全球的视频芯片供应商不反过来,让游戏商品化呢?因为这样做要难得多。 如果游戏《光环》卖疯了的话,其实是没有任何替代品的。 你不会因为去电影院看《星球大战:克隆人的进攻》太贵而决定干脆换伍迪·艾伦的电影看看。 那两部可能都是很棒的电影,但相互间并非完美的替代品。 那么:你是更愿意成为游戏发行商还是视频芯片供应商呢?对你的互补品进行商品化理解这一策略,实际上就能够很好地解释为什么很多商业公司愿意为开源做出巨大贡献。下面我们就来看看。头条:   IBM投入数百万美元用于开发开源软件 。迷思:他们这样做,是因为郭士纳看了GNU宣言后,悟出其实自己并不喜欢资本主义。现实:他们这样做是因为IBM正在成为一家IT咨询公司。 IT咨询是企业软件的互补品。 因此,IBM需要将 企业软件 商品化 ,而最好的方法就是支持开源。 你看吧,他们的咨询部门凭借着这一战略已经获利颇丰。头条:Netscape开源了他们的Web浏览器。迷思:他们这样做是为了获得坐在新西兰网吧的人提供的免费的源代码贡献  。现实:他们这样做是为了把web浏览器商品化。从第一天开始这就是Netscape的战略 。 不妨看看Netscape的第一份新闻稿 :浏览器是“免费软件”。Netscape放弃了浏览器,好让自己可以在服务器上赚钱。 浏览器和服务器是典型的互补品。 浏览器越便宜,卖出的服务器就越多。 这一点从来没有像1994年10月那样体现得那么真切。后来Netscape之所以以开源的形式发布Mozilla,是因为他们看到了降低浏览器开发成本的机会。 这样一来,他们可以以更低的成本获得商品化的好处。后来美国在线/时代华纳收购了Netscape。 但本应是商品化浏览器受益者的服务器软件做得不是太好,然后被抛弃了。 可现在为什么美国在线/时代华纳还在继续往开源投入资金?美国在线/时代华纳是一家娱乐公司。 娱乐公司是包括Web浏览器在内的各种类型的娱乐交付平台的互补品。 这个庞大的企业集团的战略利益就是让娱乐的交付——web浏览器——成为一种没人可以去收钱的商品。我的观点在Internet Explorer这里有点受挫,因为IE是免费软件。 微软也希望将Web浏览器变成商品,这样他们就可以卖桌面和服务器操作系统。 后来他们更进一步,提供了一系列的组件,任何人都可以利用这些组件组合出一个Web浏览器。 Neoplanet、AOL和Juno就用了这些组件来组建自己的Web浏览器。 既然IE已经是免费的,Netscape让浏览器“更便宜”的动机是什么? 为了先发制人。 他们需要阻止微软对Web浏览器,哪怕是免费的web浏览器的完全垄断,因为理论上这会让微软有机会以其他方式增加web浏览的成本——比方说,提高Windows的价格。(考虑到在Barksdale治下的Netscape不太知道自己在做什么,我的观点就更加不可靠了。A more likely explanation for what Netscape was doing at the time was that their top management was technically incompetent and had no choice but to adopt any solution proposed by the developer. But developers are hackers, not economists, who just happen to come up with a plan that can serve their strategy. But let’s be suspicious. Headline: Transmeta hired Linus to let him make a big impact on Linux. Myth: They do it just for publicity. Otherwise, have you heard of Transmeta? Reality: Transmeta is a CPU company. The natural complement of the CPU is the operating system. Transmeta wants the operating system to be a commodity. Headlines: Sun and HP spend money to make Ximian transform Gnome. Myth: Sun and HP support free software because they like the market rather than the cathedral. (Editor’s Note: “Cathedral and Market” is a book on software engineering methodology written by Eric Steven Raymond. The cathedral and market represent two different free software development models.) Reality: Sun and HP are hardware companies. They made boxes. In order to make money on the desktop, they need a window system, which is a complement to desktop computers, and they hope this can be a commodity. So why don’t they pay for Ximan’s money to develop their own proprietary Windows system? They tried (Sun has NeWS, Hewlett-Packard has New Wave), but these two are essentially hardware companies, so the software skills are rough, but the window system needs to be a cheap commodity, not a money that they must spend. Exclusive advantage. So they hired Ximian’s talent to do this for the same reason that Sun acquired Star Office and opened it up: commoditize the software and then make more money on the hardware. Headline: Sun develops Java; the new “Bytecode” system means writing once and running anywhere. The idea of ​​bytecode is not new – programmers always try to get their code to run on as many machines as possible. (You are so commoditizing complementary products). For years, Microsoft has had its own pseudo-code compiler and portable window layer that allows Excel to run on Mac, Windows, OS/2, and Motorola, Intel, Alpha, MIPS, and PowerPC chips. Quark also has a layer that can run Macintosh code on Windows. The best description of the C language is the assembly language of the independent hardware. This is not a new idea for software developers. If you can run your software anywhere, then the hardware is more like a commodity. As hardware prices continue to fall and the market expands, it will drive an increase in demand for software (and also allow customers to spend more on software, so software is getting more expensive now.) So Sun said to WORA (once It’s a bit strange to write the enthusiasm of running anywhere, because Sun is a hardware company. Commercializing hardware should be the last thing they want to do. Oops! Sun is a wonderful work in the computer industry, and he likes to go his own way, regardless of the consequences. Because they can’t face their past fears and dislikes of Microsoft, the strategy they adopt is out of anger rather than for their own benefit. Sun’s two strategies are (a) commercializing software by promoting and developing free software (Star Office, Linux, Apache, Gnome, etc.), and (b) making hardware by promoting Java and its bytecode architecture and WORA. Commercialization. Ok, Sun, the surprise test came: Where is your seat when the music stops? Without the proprietary advantages of hardware or software, you have to accept commercial prices, but almost the cost of a cheap factory in Guadalajara can’t be supported, let alone affordable housing in Silicon Valley. . Jared said: “And slow Joel! Just like Transmeta, Sun is trying to commercialize the operating system, not the hardware.”   也许吧,但事实上Java字节码也让硬件商品化了,这种严重的附带伤害是难以为继的。上述例子还有一件重要事情需要注意,软件对硬件商品化很容易(你只需要编写一些硬件抽象层,比如像Windows NT的HAL,只占一小块代码),但硬件想要对软件进行商品化难度就很大。 软件不是可互换的,StarOffice的营销团队正在了解到这一点。 甚至哪怕价格为零时,从Microsoft Office切换过来的成本也不为零。 在转换成本变为零之前,桌面办公软件都不是真正的商品。 即便是最小的差异也会使两个软件包之间难以切换。 尽管Mozilla拥有我想要的所有功能,只要能避免那种打鼹鼠式的弹出广告游戏的话我也很乐意使用它,但是我太习惯按Alt + D转到地址栏了。 所以你告我吧。 一个小小的差异你会失去商品化地位。 但我已经把 IBM计算机的硬盘拔出来插到戴尔的计算机上了,而且系统跑得非常好,就好像它还是在旧的计算机里面一样。Creo CEO Amos Michelson告诉我,他的公司要求每一位员工都必须学习他所谓的“经济思维”的课程。很棒的想法。 即便是基本微观经济学里面的简单概念也能充分解释今天正在发生的一些根本性转变。原文链接:https://www.joelonsoftware.com/2002/06/12/strategy-letter-v/译者:boxi。。

Tags : Internet entrepreneurship互联网创业项目创业