The Changing Software Business: Moving from Products to Services
Michael A.Cusumano
Massachusetts Institute of Technology
软件业的变化趋势：从制作转向服务
迈克尔·库苏马诺
麻省理工学院
A dramatic shift is under way in the enterprise-software industry as established vendors embrace services in the wake of declining product revenues. It remains to be seen whether life-cycle dynamics or business-model choices are behind the long-term trend.
一个引人注目的转变正在企业软件产业中进行，企业软件产业作为固定供应商，它的服务随着产品价格的下降而下降。这种长期趋势的背后，是生命周期的动态变化还是商业模式的选择，仍有侍观察。
The dramatic changes in the software business over the past few years have important implications for both users and producers of software products and services. Traditional product sales and license fees have declined, and product company revenues have shifted to services such as annual maintenance payments that entitle users to patches, minor upgrades, and often technical support.
这种软件产业的变化趋势，在过去几年对使用软件的用户和为用户提供软件产品和软件售后服务的生产者都具有重要的影响。传统产品的销售费用和牌照费用有所下降，产品生产厂家的收入都移到服务方面上，如每年的赡养费，为用户提供产品补丁，次要升级以及平时的一些技术支持。
This shift has been especially pronounced among enterprise-software vendors. We can clearly see this in the case of Siebel, whose product sales fell dramatically before Oracle acquired the company in 2005. A decade ago, even Oracle experienced the crisscross—service and maintenance revenues crossing over to exceed product revenues. We couldn’t tell if Oracle and Siebel’s product sales were dropping or product prices were falling, as Figure 1 depicts, but the effect was the same: Services (including maintenance, which typically accounts for up to 60 percent of service revenues) became more important than product revenues.
这种转变在企业软件供应商之间已经显得格外突出。我们可以清楚地看到Siebel公司的
处境，Siebel公司的销售额急剧下降，在2005年甲骨文收购其之后才有所好转。在十年前，实际上甲骨文就经历了服务和维护收入与产品收入交替超越的情况。我们不能片面的认为这种情况是由甲骨文和Siebel公司的产品销售额下降或是产品出售价格下降造成的，正如图1所示，就算产品销售额和产品出售价格不下降，造成的效果也是一样：即服务收入（包括维修，约占服务收入的60%）将比产品收入更加重要。
`There are some exceptions. Products sales continue to account for most of game-software revenues, although onlinr-gaming service revenues are growing fast. Platform companies like Microsoft—which has a large ecosystem of PC manufactures as well as enterprise and individual users driving sales of Windows and Office—continue to generate enormous revenues from products. But even Microsoft is encountering change. The company reported that services in the server and tools segment accounted for about 3 percent of its fiscal year 2007 revenues and online services (MSN) for 5 percent of its revenues. Just a few years ago, Microsoft derived all its revenues from product sales.
也有一些例外。虽然目前online游戏服务收入正快速增长，但产品销售收入仍占游戏软件公司总收入的大多数。如微软这样拥有一个大型的个人电脑制造系统的公司也向企业和特别用户推荐Windows视窗操作系统和Office系列办公软件，微软继续从产品中获得巨大的收入。但即使是遇到了变化，微软在其2007年财务报表中服务器和工具部门的收入占到了5%、在线服务(MSN)的收入占到了3%。可是就在短短几年前，微软所有的收入都来自产品销售。
A LONG-TERM TREND
一个长期的趋势
Services’ growing importance for software product firms dates back to at least 1990. The advent of free and open source software (which drove down software prices), as well as Y2K and the Internet boom and bust, accelerated the trend. In general, since 2000 or so, we’ve seen many enterprises and individual customers rebel against paying a lot of money for standardized or commodity-type software products.
服务费用的不断增长对企业的重要性可以追溯到1990年以前。自由软件和开源软件(压低软件价格)，到2000年互联网的繁荣和萧条让这种愈演愈烈。总之，自2000年左右，我们看到许多企业和特殊用户拒绝为标准软件和商用软件支付过多的钱。
New pricing models
新的定价模式
A complicating factor is the rise of new business and pricing models such as software as a service (SaaS) and “free, but not free” software. Companies like Google, Yahoo!, and even Microsoft (with Windows Live and Office Live) now deliver what used to be for-fee software products ranging from search and e-mail to basic desktop applications as a nominally free service. The user doesn’t directly pay for the software (unless you count the time to watch advertisements), but advertisers pay the software service vendor.
新业务的兴起和定价模式是一个极为复杂的因素，如软件即服务（简称SaaS）和“是免费，但又不是免费的”软件。像谷歌、雅虎，甚至微软（包含Windows Live 和Office Live服务）目前把以前的一些需要收费的软件产品作为名义上的免费服务软件，这些软件主要是搜索引擎和E-Mail等基本的桌面应用程序软件产品。用户不用再为使用软件支付费用（除非你不想看广告），但是广告商要为这种软件服务支付主定的费用。
SaaS vendors such as Salesforce. com still count SaaS as product revenues , and keep them separte from professional services. However, the SaaS pricing model actually eliminates maintenance payments—a major source of service revenues for software companies—and often includes some bundled technical support—a source of costs. So the SaaS model has confused the traditional separation of product and service revenues as well as costs, and this should result in a decline in service revenues because of the elimination of maintenance payments’.(2)
如Saleforce.com这样的SaaS供应商仍然把SaaS作为产品收入，并且保持他们独特的专业服务。然而，SaaS的定价模式实际上排除了维修费用，他们的产品通常会捆绑一些技术支持。因此，不同于传统的SaaS模式导致服务收入下降，并且最终可能导致的结果是消除维修费用导致服务收入下降。（2）
Life cycle or business choice
是生命周期还是商业选择
What’s happening to software product companies, especially those selling to enterprise customers, might be either a consequence of their lift cycles or a business model choice to emphasize services more than product sales. The lift-cycle idea suggests that software product companies start out generating most of their revenues from product license fees, but over time shift to a mixture of products and services and eventually to mostly services.
软件产品公司，特别是为企业用户提供服务的公司，他们会对软件生命周期或商业模式进行决择，相比产品而言他们更强调服务。升降周期理论建议，软件产品公司从开始出售产品时收取大部分的执照费，之后随着时间的推移，总收入转变为产品和服务的混合收入，最终以服务收入为主。
Firms might want to continue focusing on products because they can generate up to 99 percent gross margins, given that the marginal cost is zero to copy a piece of software or any other digital product. By contrast, margins for labor-intensive IT services can be 30 percent or lower.
而有的厂商可能要把产品放在首位，因为它们可以从产品上获得99%的毛利率，对于可以复制的数字型的软件生产成本基本是零。相比之下，劳动力密集的IT服务行业的利润率只有不到30%。
As competitors appear, software product companies have trouble getting new customers, or are forced to lower prices due to competition from similar firms or free software. Then these companies are more subject to what I call the”99 percent of zero is zero” rule: The great profit opportunity from software products becomes theoretical and not practical. And, whether they like it or not, their revenues gradually shift to services.
随着竞争对手的出现，软件产品公司很难开发新客户，原因是类似的公司太多和自由软件的竞争。然后，这些公司更受我之前所说的“零的99%是零”原则：从软件产品的巨大利润机会的理论变为不实际。因此，不管他们喜欢与否，他们的收入都逐步转向服务。
There’s more going on here than either an inevitable lift-cycle effect or, in some cases, explicit managerial decisions to emphasize services more than products. On the one hand, if we look at other industries, usually in the beginning of their histories, we see a lot of attention paid to product innovation and design. Once companies get the product designs right or a dominant design emerges, they shift their emphasis to the process side, such as mass production, in a product-process lift cycle.(3)
生命周期正常进行所产生的效果是不可避免的，在某些情况下，明确的强调服务比产品更重要。一方面，如果我们观察一下其它行业，通常在创业初期，它们十分重视产品的创新和设计。一旦企业获得产品的设计权或主导权，他们就会将重点转向过程方面，如大量生产产品的过程中所产生的周期问题。（3）
Striving for efficiency
Firms aim for production efficiencies. In the early 1900s, Ford introduced the Model T(which became the standard automobile design), then focused on standardizing components and automating mass production. In the software industry, these’s been a shift from product design in the 1960s to software engineering in the 1970s and 1980s, culminating in “software factories” in Japan and India, as well as the Capability Maturity Model in the US.
追求效率
企业的目标是生产效率。在19世纪初期，福特引进了T模型（成为汽车设计的标准），之后便集中精力在于部件的批量生产的规范和自动化。在软件产业中，产品效率的改变是从1960年代逐渐发展到1970年代和1980年代的软件工程，最后日本和印度成为“软件工厂”，成熟的软件模型在美国。
Service innovation is an aspect of the lift cycle that might affect software and some other industries. For example, if the product design has become a commodity—widely available and low-priced around the world with little differentiation — and after a company has wrung maximum efficiency out of process improvement—then management might turn its attention to services.
服务创新可能是影响生命周期软件和其它行业的一个方面。举例来说，如果产品设计已经成为一种日用品普及，价格低，小公司心经不能通过过程改进来提高最大生产效率的可能，因此管理者的注意力都将转向服务。
On the other hand, what we’re seeing might be related to “S-curves” and “disruptive
technologies.” (4) In software, not only do we have maturity setting in for different product segments and companies shifting their emphasis to services, but some new technologies now support different kinds of business models, including different ways of pricing and delivering software, and reaching different kinds of customers.
另一方面，我们正在分析“S曲线”和“突破性的技术”之间的联系。（4）在软件方面，
我们的产品不仅有不同的分类用来满足强调服务的公司，而且现在一些新的技术支持不同各类的商业模式，包括不同的定价和自定义付费软件以满足不同的客户。
Obviously the Internet and wireless technologies enable all sorts of on-demand or transaction-based pricing models or Google types of advertising-based revenue models, In addition, a platform transition seems to generate demand not only for buying new products but also for services. For example, a customer switching platforms form mainframe to client-server or form client server to the Internet or from stationary to mobile probably needs a lot of services in terms of strategic assistance, rewriting applications and data, or retraining employees. In other words, platform transitions such as we’ve experienced over the past 15 years could also generate as much or more new revenue from services as from products, especially since many products are now free or low-priced.
显然，互联网和无线技术能够满足各种交易和需求以及谷歌基于广告的定价收费模式，除此之外，一个平台的过渡不仅需要新产品而且需要服务。例如，一个客户改变平台从主机到服务器，从客户服务器到互联网，从静止到移动可能需要很多的服务策略，如修改应用程序和数据或者再培训员工。换句话说，自15年前我们经历平台转变以来，已经产生了尽可能多的产品服务收入，特别是现在很多产品是免费或低价的。
SIMULTANEOUS MATURITY AND INNOVATION
成熟与创新同步
To sort out what’s happening in the software business, I launched a research project at MIT in 2003 to examine this shift from products to services for companies in software and other industries. My colleagues Fernando Suarez and Steven Kahl and I are still analyzing the data, but we have some preliminary findings and observations.
归纳整理了目前发生的软件业务，我于2003年在麻省理工学院建立了一个研究项目，用来研究软件公司和其他行业从产品到服务的转变。我的同事Fernando Suarez和Steven Kahl
一直分析这些数据，并且我们有了一些初步的调查结果和观点。
Peak and consolidation
The first database we created, covering 1990 through 2003, is a comprehensive list of firms that consider themselves software product companies selling “prepackaged software,” listed under US Standard Industrial Classification (SIC) code 7372, as Figure 2a illustrates. This data includes foreign firms such as SAP and Business Objects that list on US stock exchanges, as well as game-software firms that sell products almost exclusively.
我们创造的第一个数据库，包括1990年到2003年，是一个公司的综合列表，这个综合列表中的公司认为自己是软件产品公司，并且是销售“预先包装好的软件，”登记在美国标准工业分类（SIC）编号7372中，好图2a所示。这些数据包括外国公司，如SAP与美国证券交易所上的商业对象，也包括几乎所有的销售游戏软件的公司。
The data set contains about 500 distinct firms and peaked in 1997 at about 400 firms. By 2006, the list was down to fewer than 150 firms—indicating a dramatic consolidation of the software products business.
该数据集合包含了大约500个截然不同的公司。到了2006年，这个名单少于了150，这标志着软件产品业务戏剧性的团结在了一起。
The second database, which covers 1990 through 2004, consists of firms that compete in IT services under several different SIC codes. This data, illustrated in Figure 2b, also shows listed companies peaking in 1999 at just below 500, and declining to less than 250 in 2004. The strong rise in IT services companies in the 1990s suggests that the transition from client server to Internet platforms provided as many or more opportunities for services firms as it did for software product firms to become public companies, though both the services and products side of the business have experienced significant consolidation since that time.
第二个数据库是从1990年到2004年，包含了一些不同SIC编码的IT服务公司。这个数据如图2b所示，也可以说明上市公司在1999年达到了500家的顶峰，到了2004年下降至250家。90年代大幅上升的IT服务公司表明，从客户服务器向互联网过渡是服务公司取代产品公司成为上市公司的机会。从那时起服务业和产品制造业进行了融合。
The fact that the number of public software and IT services companies is consolidating suggests that the software business is maturing. However, other data collected at MIT suggests a strong rise in start-up enterprise-software companies, especially in 2005, using a variety of new business or pricing models (www.agoeldi.com/media/Thesis_AGoeldi_Final_09MAY07.pdf).
事实上，产品和服务的联合表明软件业务逐步走向成熟。然而，在麻省理工学院收集的其他数据显示在2005年软件公司大量上升，使用新的业务或定价模型。
New business models
新的商业模式
Figure 3 shows the business models of 108 companies competing in Web-based enterprise software (about 20 percent of the companies are publicly listed), and indicates that monthly subscription fees are the most popular pricing model. A minority of companies also offered free software or advertising-based software (Google falls into this category), and others charged the traditional license fee.
图3显示108家基于网络软件商业模式的公司（大约20%的上市公司）竞争市场，并且一致指出按月收费是最受欢迎的定价模式。少数公司也提供免费软件或者基于广告的软件（谷歌掉进这一类），用其他的收入填补执照费。
Figure 4 shows a model my MIT students made that categorizes the variations now occurring in revenue of business models, delivery models, and target customers. A decade ago, nearly all software through the up-front license fee and did local installations on the customers’ hardware. Now we have many different business models—subscription, advertising-based, transaction-based, and several kinds of” free, but not free.”
图4是我在宁可省理工学院的一个学生制作的，他将商业模式，交货模式，及客户群
的收入变化进行分类。十年前，几乎所有的软件都是收取前期执照费，并且把软件安装在客户的主机上。现在我们有不同的商业模式进行选择，如订阅的，基于广告的，基于交易的和“免费，但又不是免费”的服务类型。
Software delivery models can be remote and web-based or bundled as hardware products. This trend toward potentially cheaper software, combined with less costly ways of delivering software over the Web, has made it possible for firms to target not only mainstream customers but small businesses and leading-edge early adopters.
软件交付模式可以使用远程方式和基于网络的捆绑硬件产品方式。廉价的软件是一个潜在的趋势，结合软件方式提供和网站低价的特点，可能不仅将目标指向大型企业而且将小企业发展成主流客户。
In addition, many software companies are now turning into hardware companies in what’s sometimes called the “appliance model” (http://dspace.mit.edu/handle/1721.1/39504). If you put the software in a box, it’s less likely that the price will fall to zero. People usually will pay more for a box, even though they might not want to pay much for software or digital media on its own.
此外，许多软件公司现在变成硬件公司，有时也被称为“家电模型”。如果你把软件放在一个盒子里，并把它的价格降为零。人们通常可能会花更多的钱去买这个盒子，尽管他们或许不想去买盒子里的软件，或是盒子里的数字媒体等软件产品。
Another element behind this entrepreneurial activity is that it might take less money to start a software company. Of course, it was always possible for “two guys in a garage” to launch a software or computer-hardware company, and many started that way. But today, many critical enterprise components—the operating system, database, and web applications server—are available as free and open source software. An entrepreneur can write some applications code and then hire another firm to host the software and, with relatively little expense, launch an enterprise-software company. Data from a recent survey suggests that entrepreneurs funded about 37 percent of the new web-based enterprise start-ups, and only 36 percent relied on venture capital (www.agoeldi.com/media/Thesis_AGoeldi_Final_09MAY07.pdf).
另外一个因素是中间商可能会花很少的钱创立一家软件公司。当然,许多时候这种两个人一起合作创立一个软件公司或硬件公司是最有可能的。但目前,许多重要的操作系统、数据库和网络应用程序服务器的企业部件都是很有用的开源软件。企业家可以写一些应用程序代码,然后雇用另一家公司以相当少的费用,创建一个企业软件公司。从最近的调查数据表明,新式网络企业家初创企业,仅有36%依靠创业资本。
Temporary or permanent?
是临时还是永久?
As we look back at these trends and new developments in the software products business, a question occurs: Is this increase in services and new business models temporary or permanent? Permanent in my mind refers not necessarily to “forever” but to a trend lasting decades rather than years.
当我们回顾这些趋势和软件产品的新型发展,一个问题出现了:这是增长的服务和新商业模式是临时性的还是永久性的?在我的脑海里指的不是永久或是一定要“永远”,而是一个趋势持续几十年而不是几年。
One possibility is that we’re now merely in between platform transitions and probably at a bit of a plateau in terms of product revenue growth. If some major innovation occurs, such as for a new computing platform, then individuals and enterprises will again start buying new products, both hardware and software, in large numbers.
一种可能性是我们现在只是在不同的平台上转换，也可能产品收入正在稳定增长。如果一些重要的改革发生,比如一个新的计算平台,那么众多的个人和企业将再次重新买入包括硬件和软件的产品。
By contrast, the permanent argument says that software might have experienced what computer hardware did in the past: Investments from Intel and other firms along the lines of Moore’s law helped dramatically reduce the price of computing power and bring powerful computers down to the level of commodities.
相比之下, 永久争论软件说可能会经历以前计算机硬件所经历的那样:英特尔与其它企业沿摩尔定律的路线发展，大幅降低高质量计算机的价格的同时也带来了计算能力和水平低下的商品。
In other words, the permanent argument suggests that much software now is also commoditized, just like hardware, and prices will fall to zero or near zero for any kind of standardized product. In this scenario, the future is really free software, inexpensive SaaS, or “free, but not free” software, with some kind of indirect pricing model, like advertising—a Google-type of model. And it’s possible that other commoditized high-tech industries, especially those with significant value coming from software, are likely to follow.
换句话说，永久争论表明现在很多软件仍然是商品化的,就像硬件, 对于任何一种标准化的产品，价格会降至零或接近零。在这种情况下,免费软件是真正的未来,针对SaaS,或“免费，但又不是免费”的软件，
带着某种间接定价模型,就像谷歌的广告模式一样。其他可能的,就可能出现高科技商品产业,尤其是那些有显著的价值的软件将会接踵而至。
WHAT THE DATA SAYS
Perhaps we’ve experienced changes that are long-term, rather than temporary. But what does the data say? Our database of 500 publicly listed software product companies contains an average of about 10 years of data for each firm (totaling over 3,200 annual observations). Excluding game-software firms and some other firms (mostly, they didn’t break out products versus services and we couldn’t confidently classify their revenues), the total number of firms peaked at 300 in 1997 and stood at merely 111 in 2006. As Figure 5 shows, software product firms in our sample had an average of 70 percent of their revenues coming from product sales in 1990 and less than 50 percent since around 2003, when the crisscross first happened for the industry as a whole. If we remove game-software companies from the sample, the crisscross happened in 2002 and is a bit more pronounced.
什么是数据论
也许我们曾经经历改变的影响是长期的而不是短期的。但是数据说明了什么?我们大约500家的数据库产品上市公司 平均为每个公司存储大约10年的数据(总计超过3200年的观察)。不包含游戏软件公司和其他公司(他们通常不打破与产品与服务的原则,我们不能明确的归入这个分类),在1997年他们的总数达到300家的巅峰,而到2006年他们只有111家。如图5所示, 在1990年，软件产品厂商平均有70%的收入来自产品的销售, 在2003年则不到50%,当第一次发生的交错后整个产业就形成一个整体。如果我们排除游戏软件公司, 第一次发生的交错将在2002年,这点就更加明显。
We didn’t separate maintenance from other services because less than 10 percent of our sample broke this our. Firms treat maintenance as a type of service because, unlike with product sales, companies can recognize these revenues only as they deliver patches, upgrades, or technical support over time.
我们没有将维修从服务中分离出来,因为只有不到10%的公司打破了常规。公司把维护作为服务的一个类型,因为与产品销售不同,公司可以识别出这些收入当他们提供升级补丁、升级、或技术支持的时候。 Some firms, such as SAP and Oracle, are now trying to relabel maintenance fees as product revenues in the sense that they represent product renewals. This makes some sense because maintenance has profit margins closer to product sales (though a bit lower because of the routine technical support costs usually included in the maintenance agreements), but maintenance revenues and recognized over time, like other services.
一些公司,比如SAP和Oracle,正在努力确认维护费用在某种意义上作为产品收入,因为他们代表产品更新。这使得有些感觉,因为维持产品的利润接近销售产品的利润(虽然略低,因为日常技术支持的成本通常包含在维护成本中),但随着时间的延长维护费用就会象其他的服务那样被认可。
Reaching equilibrium
The data indicates that product revenues have dropped but haven’t continued to fall to zero. Rather, they’ve stabilized at just over 50 percent of total revenues. So perhaps software product companies have reached a sort of equilibrium point as a business—more service (including maintenance) revenues from their existing customers than new-product revenues, but products are still holding significant value, at least for the publicly traded companies, we see this stabilization trend.
达到平衡
数据表明,该产品收入下降了,但不会一直下降至零。相反,他们已经稳定在刚刚超过50%的总收入。因此软件产品公司的现有客户的商业服务(包括维修)收入与他们的新产品的收入或许已经达到一种平衡,而且产品仍然保持着显著的价值,至少在公开上市的公司中,我们看到了这个稳定的趋势。
We can also look at how common it is for software product companies to sell only products as well as have different hybrid mixtures of products and services. In 1995, Richard Selby and I published a book that held Microsoft up as the ideal model for a software company—100 percent product revenues and those wonderful gross margins.(5) But the data suggests that these kinds of companies are relatively rare historically and are becoming fewer over time.
我们也可以看看把常见的软件产品公司中只卖产品与混合物销售产品和服务相对比。1995年,理查德·塞尔比和我出版了一本书,展示了微软的理想模式,一个软件公司能在产品上获得100%的毛利润。(5)但数据表明,这些类型的公司是比较少见的历史和变得更少。
Our preliminary analysis also indicates that, while the average level of product revenues has dropped to less than 50 percent for the software product companies, the optimum mix for operating profitability (again, excluding games and some other firms) seems to be at about 70 percent products and 30 percent services. There are also some companies in our database that have reported 100 percent service (including maintenance) revenues in a given year and no product sales, even though they’re nominally software product companies. Companies in this category are likely to be weak performers and candidates for takeover or bankruptcy.
我们的初步分析也表明,当产品收入已下降至不到50%的软件产品公司,经营利润的最优组合(同样,不含游戏和其他一些公司)似乎是70%左右的产品和30%的服务。也有一些公司在我们的数据库中,也有一些名义上的软件产品公司报导称拥有100%的服务(包括维修)收入,并且没有产品销售。这种类型的公司极有可能是核心管理人员大调整或是企业濒临破产。
Reasons for the shift
Why the shift toward services? On the surface, primarily it’s happening because software product firms are getting older. They creep toward that service crisscross at the rate of nearly 2 percent a year. The crisscross point by age is at 26 years for the whole sample and 22 years if we exclude game companies. In other words, if a software product company survives for more than 20 years (and doesn’t sell software games), it’s likely that service and maintenance revenues now equal or exceed product revenues.
转变的原因
为什么会向服务转变呢?表面上看,它的发生主要上由于软件产品公司的老化。他们由产品转向服务的增长速度大约在2%每年。产品收入和服务收入的交错点在26年，如果我们排除游戏公司那么这个交错点将会是22年。换句话说,如果一个软件产品企业规模超过20年(并且不卖游戏软体),很有可能,这个企业现在服务和维护的费用就会等于或超过产品收入。
When we probe more deeply, statistical regression analyses suggest that this transition is also related to lagging growth in product sales and total sales, as well as the recession that followed the Internet boom. The appearance of the Internet as a disruptive new platform also generated new service sales, especially for IT services companies. But this factor is statistically less important than firm-level factors for the product firms, such as age and lag in sales.
当我们更深入的调查和统计回归分析后发现,这一转变也与产品销售增长滞后,总销售量以及网路经济衰退有关。国际互联网平台的出现破坏产品的同时也催生新的服务销售,尤其是对IT服务公司而言。但相比年龄问题和销售的缓慢增问题对于统计的一级产品公司，这一因素是次要的。
In other words, the shift toward services for product firms appears to have happened for two reasons. One is that product sales might continue to grow, but services grow faster, perhaps because price levels or the number of new customers falls. This situation is still relatively healthy, and firms can easily survive as hybrid businesses. The other scenario is that the products business collapses, and that’s why firms cross over to a majority of service revenues.
换句话说,导致产品公司向服务转变的原因只有两个。一是产品的销售可能会继续增长,但服务会更快地增长,这或许与物价水平或新客户的数量下降有关。这种情况下相对来说还是健康的,企业可以很容易地在将业务混合销售。另一种是产品业崩溃,这就是为什么大部分公司偏重服务收入。
This second scenario is potentially disastrous because it often means the firm must reorganize radically and perhaps quickly, as in the case of Siebel or another firm I’ve written about, i2 Technologies.(6) The firm can no longer support large product R&D groups with large marketing and sales expenditures. It must transition from designing products for a largely abstract set of users to building and servicing products for individual customers. Many firms don’t make this transition or make it poorly and reluctantly, missing the opportunity to manage services as a strategic area.
这第二个原因是存在潜在灾难性的,因为它通常意味着公司必须尽可能快的彻底重组, Siebel公司的案例就是一个最好的例子。(6)公司不再投入大量金钱去支持大型产品的研发团队与大市场和销售。它必须在确定的用户的需求后再做出转变以应对客户，并且在以后才能为客户做出更好的服务。许多公司不会做任何转变或不情愿地做出转变,他们错过了一个战略性的向管理服务转变的机会。
IMPLICATIONS FOR PERFORMANCE AND STRATEGY
As we collected our data, my research colleagues and I thought the impact of rising services would have a negative impact on profitability and market value for a software product company because services tend to have lower profit margins and signal lower growth prospects. What we’re seeing, however, is a more complex relationship.
性能和策略的影响
当收集资料时,我和我的同事们研究认为服务会上升对一个软件产品的公司的盈利能力和市场价值负面影响,因为这是服务走向低利润、产品低增长的趋势。这只上我们所能看到的,这中间其实有更复杂的关系。
For most software product companies, services generally contribute positively to their profits, but not in the linear manner we’d expected. More specifically, there seem to be “sweet spots” at the low and high ends of the spectrum. We can roughly say that, for the average software product company (excluding game software), services contribute positively to profits until they account for about 20 percent of total revenues.
对于大多数软件产品公司,服务通常支撑着他们的利润,而不是我们所预测的那样。更确切地说,在那里似乎是区别低端和高端范围的一个“最有效点”。我们可以大致说,对于一般的软件产品公司(不包括游戏软件公司),服务收入利润稳步增长，直到占总收入的20%。
After that point, services become a drag on profitability until they reach about 60 percent of revenues. Then services begin again to have a positive impact. One possible explanation for this curvilinear effect is that product companies might sign most of their customers to simple maintenance contracts for up to 20 percent or so of the retail price of their products, and these kinds of services are very profitable for them as long as technical support costs are minimal.
那之后,服务已经影响到它们大约60%的趎利润收入。因此服务重新开始产生重大影响。一种对曲线效最有可能的解释就是,该公司大部分占产品零售价格20%左右的维护合同, 可能会丧失了他们的顾客，因为这些类型的服务只要技术支持的成本较低都将有利可图。
But as the product companies get deeper into labor-intensive services, such as product customization and complex integration work, or strategic consulting and training, services can become a drain on profits until the product companies gain enough scale and experience to perform these services efficiently. Then they begin again to make money from services, much like dedicated IT services companies do. SAP and Oracle would fit this model. Both are very profitable and have only about one-third of their revenues coming from new product license sales.
但当产品公司更进一步深入到劳动密集型服务时,如产品客户化、复杂的整合工作、战略性的咨询和培训，服务直到获得足够的规模和经验来执行这些服务效率时长能成为产品公司的利润点。然后他们开始从服务赚钱,就像SAP和Oracle这样适合该模型的专用IT服务公司一样。 两种都很有利可图并且只有大约三分之一的收入来自新产品许可证销售。
Market value, which generally tracks growth rates as well as profitability, follows a similar nonlinear curve. It seems to be positive until about 20 percent, then negative until about 80 percent, and then positive again. However, our data also indicates that, even in years when services positively contribute to profitability, market cap can drop as services rise. This suggests that investors still don’t understand how important services have become to the revenues and profitability of software product companies.
市场价值通常随成长率的增长而增长,就像一个类似非线性的曲线一样。这看似是积极的,其实只有大约20%,那么还有80%是消极的。然而,我们的数据也显示,在几年内，市场价值可降低服务增长。这表明投资者还是不理解服务对软件产品公司的收入和利润有多重要。
Services as a strategic area
The positive impact of services on profitability and market value differs somewhat by product category, and we’re still in the process of sorting out these differences. But the general conclusion seems to be that many or most software product firms can and should take advantage of services, especially maintenance, and not just let services “happen” because their product business declines.
服务战略区域
这对不同产品类别服务的盈利和市场价值有积极的作用,我们仍然在不断的整理这些差异。但一般的结论似乎是许多或大多数软件产品企业从服务中获利,特别是维护,而不仅仅是服务使他们的产品销售有所下降。
This means that software product firms—and probably many other high-technology firms—should treat services as a strategic area and a target of opportunity to increase and profits—especially when the product business is suffering. We can see this in another preliminary analysis which suggests that, for every 10 percent increase in maintenance as a percentage of total services, service gross margins rise about 5 percent. In other words, if the products business is declining and shifting to services, companies should try to sign every customer to a maintenance agreement to minimize the impact on profitability.
这意味着软件产品公司和很多其他高科技公司应该探讨一个服务的目标战略,用来增加利润尤其是当产品业务萧条。我们可以看到这在另一个初步分析表明,因为在维护整体服务上每增加10%的百分比，服务的毛利率就将上升约5%。换句话说,如果产品业务正在慢慢减少并逐渐转向服务,公司应该尝试在每签下一个维护协议的同时，对公司盈利的影响降到最低。
By contrast, too many product firms seem to treat services as a necessary evil and manage them as a cost center, without much creativity or effort to grow that part of the business. In fact, though, most firms can look at their past trends and predict when they’ll hit the crisscross and take some strategic responses, such as trying to rejuvenate the product lineup or launching a major campaign to sell more maintenance and other services, as firms such as SAP and Oracle have done over the past decade.
对比之下,太多的产品公司似乎认为服务是一个不可避免的麻烦,并且对这部分业务不重视。事实上,大多数公司能根据过去的趋势进行预测,采取一些切实可行的战略反应,例如试图使产品焕然一新或做一些大规模的活动为公司销售更多的服务和其他维护,SAP和Oracle在过去十年都是这么做的。
We also found that this trend toward services isn’t limited to the software business, though it seems to be less of a life-cycle phenomenon and more a strategic move in other closely related industries, such as computer and telecommunications hardware and equipment. For example, as Figure 7 shows, IBM’s service revenues have grown from less than 30 percent of revenues to more than 50 percent over the past decade. Sun Microsystems, Hewlett-Packard, Cisco, and even Dell have shown major increases in services and this seems to correspond to the commoditization trend in hardware.
我们发现这个趋势并非仅限于服务软件业务,虽然它看似是一个很小的生命周期现象和为发展而制定的产品战略行动,诸如计算机和硬件通信设备。例如,如图7显示,IBM的服务收入增长不到30%，在过去的十年来自服务的收入超过50%。太阳微系统公司、惠普、思科、甚至戴尔已经表示要增强服务,这似乎和硬件商品化趋势相符合。
Effect on IT services firms
The shift toward services for the product companies might be bad news for the dedicated IT services companies. Firms such as Accenture and Infosys are historically partners of enterprise-software product companies like SAP and Oracle, and they gain significant revenues by installing, integrating, and customizing enterprise systems. But services are really money that product companies “left on the table” in the hope that services partners can help them sell more products. If the product revenues disappear, however, then former partners must compete for the same money.
IT服务公司的影响
产品公司向服务转变可能对致力于IT服务的公司是个坏消息。如Accenture和Infosys这样的合作伙伴,就像SAP和Oracle一样,他们从安装、整合、客户化企业系统中获取显著的收入。但对于产品服务公司，服务代表着收入,因为他们希望服务伙伴能够帮助他们销售更多的产品。如果产品收入消失,他们会通过以前的合作伙伴来获得同样的收入。
THE THREEFOLD CHALLENGE
There’s threefold challenge for managers of software product companies and other firms experiencing this shift toward services.
三重挑战
软件产品公司的管理者向服务转变要面临三重挑战。
Managing the crisscross
First, how can you manage this crisscross? Managers need to identify the best mix of product revenues (hardware and software, if appropriate) for their particular business segments along with service and maintenance revenues and determine how to impact these percentages. Services seem especially complementary in some business segments, like enterprise applications, while they’re potentially more of a drag on other segments, although recurring maintenance payments are probably good for every product company.
交错管理
首先,怎么管理这个交错?经理们需要找出最佳的组合产品收入(硬件和软件,如果适当的话)，这些百分比会根据业务领域的服务和维修随收入决定。服务似乎会在一些业务领域上互补,如企业应用,他们很可能在其他环节上的拖延,虽然循环维护可能有利于每一个产品的公司。
Another point we tend to forget is that, for most product companies, products are the engine that drives service and maintenance revenues. Products and services are coupled for most firms, even though IBM and a few other companies such as General Electric have managed to become relatively neutral vendors of services. Most product firms need to maintain strong product lineups that keep customers paying for implementation or strategic services as well as long-term maintenance contracts or subscriptions.
我们往往会忽略另一点,对于大多数产品的公司,产品含有驱动服务和维修的收入。大多数公司都使用产品和服务相结合的方式,即使IBM和其他一些如通用电气这样的公司已经成为了相对中立的服务供应商。大多数产品是公司需要维持强大的产品阵容,让顾客付费使用或者长期订阅维护服务合同。
‘ Servitizing ’ products
Second managers need to think about how they can “servitize” their products—that is,. create service offerings that add value and distinctiveness to their products. Services wrapped around products can make the products less commodity-like as well as generate new revenues and profits, even as the product business declines. In some industries, there’s evidence that services over the lifetime of the product of the product can generate several times the initial profits on the sale.(6)
服务产品
第二个是 经理需要考虑怎样服务他们的产品，并为他们独特性的产品服务增加价值。当产品业务有所下降时，产品服务需要和产品相结合来产生新的收入和利润。在一些产业中,有证据表明, 产品服务时间长的产品能够常来最初产品销售额几倍的收入。(6)
Some day soon, for example, companies will give away various devices for free and just sell services or some kind of subscription contract. The cell-phone industry is well on the way toward this path. The automobile industry might follow as well. Even today, General Motors and Ford make little or no money from their products business while nearly all their revenue comes from financial services such as loans and leasing.
很快会有一天,公司将赠送各种各样服务或出售某种订阅服务合同。手机行业在不久也将会走这条路。汽车工业也是。甚至在今天,通用汽车和福特汽车从他们的产品业务中只赚取到很少的钱甚至赚不到钱,他们几乎所有的收入都来自金融服务,如贷款和租赁。
In the automobile industry, other ecosystem players make even more money from insurance and other services. What GM and other distressed automobile companies should do is give away their products at cost and sign customers to all-inclusive lifetime services contracts—not only loans or leases, but also insurance, maintenance and repair, and telematics services like GM’s OnStar.
在汽车行业,大多数的钱来自汽车保险等服务。通用和其他汽车制造商应该像旗舰店一样按成本赠送给客户一个终生的服务合同，这个合同不仅包含贷款和租赁还应包含保险、维修、远程处理服务,。
‘Productizing’ services
Third, managers need to think about how to “productize” their services so they can deliver them more efficiently. Productization of services can come from component or design reuse, computer-aided tools, and standardized process frameworks and training, as seen in past Japanese software factories such as at Hitachi or Toshiba, or in present Indian IT services companies such as Tata Consulting Services, Wipro, and Infosys. But productization can also come through automating services, such as the way eBay, eTrade, Expedia, Google, Lending Tree, and other Internet companies deliver their software-driven products or services.
产品化服务
第三,管理者需要考虑如何产品化他们的服务,这样他们就可以提高效率。产品化的服务可以来自部件的设计重用、计算机辅助工具,标准化进程,例如过去的日本软件工厂如日立和东芝,或者像印度的塔塔咨询服务公司这样的IT服务公司。但产品化还得经历自动化服务,如易趣、谷歌,和其他的因特网公司提供他们的软件驱动产品或软件驱动服务。
In fact, fully automated services should be able to generate the same level of gross margins as a traditional software product company. That’s why web-based delivery of software that different business models support is such an intriguing change for how producers distribute, deploy, and receive payment (or don’t receive payment) for their software products and services. It’s also why Google now rivals Microsoft in profitability, market value, and leadership in the software business.
事实上,完全自动化服务应该能够产生和传统的软件产品公司同样水平的毛利率。这就是为什么基于网络交付的软件有不同的经营模式,有趣的是这种经营模式是如何为他们的软件产品和服务分配生产者、部署、和接收付款(或没有收到付款)。这也是为什么谷歌在盈利能力、市场价值和软件领导能力上和微软竞争。
In the future, as my colleagues, students, and I continue to do these kinds of analyses, we probably will change the way we think about the software business and some other high-tech sectors like Internet services, telecommunications, and digital media. There will probably always be some traditional product companies like Microsoft that package technology and sell thousands or even millions of copies of their products. But our data suggests that not only are the numbers of these companies dwindling, the survivors also have to spend a fortune on sales and advertising as well as product development. As a result, most traditional software product companies make little or no money for their investors, and that’s another reason why the smaller firms are disappearing. We would get a different picture, however, if we included companies like Google (whose SIC code lists it as an Internet services company) and perhaps some of the new SaaS start-up companies in the ranks of software product companies. Combining this data would give us a better idea of how much money customers are actually spending (directly or indirectly) on software-based products and services rather than just traditional software products—that is, including automated, standardized services and digital content delivered over the Web.
在将来,我的同事、同学和我会继续做这些类型的分析,我们可能会改变对一些高科技软件部门的思考方向,例如互联网服务、电信和网络媒体的问题。总是有一些传统的产品公司,如微软那个能卖成千上万甚至上百万份的产品包裹技术。但我们的数据表明,不仅是这些公司的数量减少,而且幸存者也要花一大笔钱用于销售和广告以及产品的开发。结果是,大多数传统软件生产公司回馈给他们的投资者很少的回报，甚至没有回报,这是另一个小公司消失的原因。这个数据给了我们一个很好的结论，顾客实际上为软件产品和服务(直接或间接) 包括自动化、标准化服务和数字内容实施网络花了多少钱。
Acknowledgments
Thanks to the following MIT students for supplying figures: Francois de Laigue for Figure 2b, Andreas Goeldi for Figure 3, and the team of Krishna Boppana, Andreas Goeldi, Bettina Hein, Paul Hsu, and Tim Jones for Figure 4.
鸣谢
感谢下列麻省理工学院的学生提供的数字: Francois de Laigue for Figure 2b, Andreas Goeldi for Figure 3, and the team of Krishna Boppana, Andreas Goeldi, Bettina Hein, Paul Hsu, and Tim Jones for Figure 4.Refernces
1. M.A.Cusumano,The Bussiness of Software, Free Press/Simon&Schuster,2004.
2. M.A.Cusumano, “The Changing Labyrinth of Software Pricing,” Comm.ACM, July 2007, pp. 19-22.
3. J.Utterback,Mastering the Dynamics of Innovation,Harvard Business School Press,1994.
4. C.M.Christensen,The Innovator’s Dilemma, Harvard. Business School Press,1997.
5. M.A.Cusumano and R.W.Selby,Microsoft Secrets,Free Press/Simon&Schuster,1995.
6. T.Knecht,R.Leszinski,and F.A.Weber, “Memo to a CEO: Making Profits after the Sale,” McKinsey Quarterly,Nov.1993, pp. 79-86.
Michael A.Cusumano is the Sloan Management Review Distinguished Professor at the Massachusetts Institute of Technology’s Sloan School of Management and Engineering Systems Division. He specializes in strategy, product development, and entrepreneurship in the computer software industry, as well as automobiles and consumer electronics. Cusumano received a PhD in history and East Asian languages from Harvard University and completed a postdoctoral fellowship in production and operations management at Harvard Business School. Contact him at cusumano@mit.edu.
本文来源：https://www.2haoxitong.net/k/doc/7fbdb17a5acfa1c7aa00ccfc.html
文档为doc格式