????克雷頓?克里斯滕森的“創新者困境”（Innovator's Dilemma）理論直至2011年的今天仍然適用，他的擁躉應該絲毫不會對此感到奇怪。顯然，創新依然很重要。要想知道創新的作用，看看你的電腦桌面就知道了。想想看，你平日里常用的桌面軟件有多少？然后把它與你和你的公司所使用的云服務的數量做一個對比。如果你也像其他大多數商務人士一樣，那么在過去的20年里，你肯定一直在與各種緩慢、痛苦、繁瑣的應用程序打交道，因為你需要它們幫你運作項目，管理客戶，共享文件。有時候軟件的安裝要花上數月的時間，有時候維護成本甚至超過了原始產品的價格。有時意想不到的延誤和宕機會導致生產率出現明顯下降。但是隨著云計算的出現，一切都改變了。

????云計算提供的企業解決方案恰好符合克里斯滕森的“破壞式創新”理論框架。它們為傳統的企業運算模式提供了更便宜、更簡便、適用范圍更廣泛的替代性選擇。它們一開始往往扮演低端的攪局者角色，靠成本和性能上的優勢來吸引已經被過度服務的客戶。不過隨著這些技術的可靠性和復雜性漸漸成熟，它們開始在企業間迅速傳播開來，并且開始解決某些企業最重要的問題——云計算的發展恰好符合這個過程。克里斯滕森一定得意地笑了。

????這一切都為傳統的企業軟件制造巨頭制造了巨大的麻煩。這些軟件巨頭也深知需要做出跨躍式的創新，但是他們缺乏創新的商業模式和DNA，也缺乏背水一戰的緊迫性。全球每年花在企業軟件上的經費超過2,700億美元，但到目前為止，這些錢只有很少的一部分裝進了云計算廠商的口袋?，F在還缺乏明確的激勵，促使傳統軟件商強勢進軍云計算領域。如果這些軟件企業突然轉型，可能會讓客戶感到困惑，擾亂渠道，公司經濟狀況失衡，或者導致執行不力。因此，困境就出現了。

為什么說云計算是一種顛覆性創新？

????當然，并非所有的創新都是顛覆性的。如果一項技術進步僅僅只是強化了現有廠商在某個特定領域的地位，那么它只能算得上是一項可持續性創新，而不是顛覆性創新。如果某項技術進步幾乎沒有給現行的業務模式或主流的市場策略帶來什么變化，或者這項技術進步只是一項非線性的技術，那么市場現有的其他參與者可以說還是安全的。例如在存儲業務領域，領先的存儲提供商（在一定范圍內）作出的性能方面的改善往往是可持續的。而全新技術的出現，比如閃存，則會使市場上的新玩家能夠利用這個機會更快地賺到錢，更快地實現新技術的商業化。可以斷言，這種情況將會不斷出現。

????互聯網剛剛出現的時候，作為回應，戴爾（Dell）公司也跟風推出了Dell.com網站，但后者對戴爾而言只是一個新渠道，賣的還是那一套產品。如果能有效利用互聯網，你會發現戴爾以前的產品目錄和它新建的網站（Dell.com）之間并沒有明顯的差別。然而，每個市場都是不同的。例如隨著整個世界向超本文方向轉變，點評網站Yelp填補了黃頁的空白，取得了成功。這并不是因為網絡本身是一項能夠連接本地信息的顛覆性創新，而是因為Yelp的用戶生成式評論提供了一種全新的方式和體驗，為用戶提供本地向導。黃頁這種產品缺乏社區概念，對當今世界毫無意義。

????當今世界，沒有什么東西是確定的，只有死亡、稅收，和企業軟件的破壞性除外。上世紀90年代，在“用戶—服務器”模式取代大型主機后，整個行業也經歷了一場劇變。1995年《紐約時報》（the New York Times）在評述仁科公司（PeopleSoft）以及用戶—服務器模式時說：“任務在‘用戶’計算機和大型主機間重新分配，并在二者間或多或少實現了平衡。”這項新的標準（當然現在已經過時了）代表了價值的劇變，因為使用了PC標準后，應用程序變得更加強大，更加現代化，數據可以大量集中，而且運行成本只相當于大型主機時代的一小部分。

????快進到15年后，這種大規模的變革再次上演。這一次，大多數的核心業務應用被搬到了網絡上。那么為什么說云計算是一次顛覆性的變革，而不是那種傳統的軟件廠商可以迅速實現的創新呢？

????這是因為云端軟件的銷售、營銷、分銷和應用等方方面面都與以前大不一樣。云服務的銷售和營銷與終端用戶的服務捆綁在一塊，而不是傳統企業軟件那種托管式的從上至下的部署。云計算的技術是通過網絡分銷的，某個市場領袖企業所獲得的技術，在本質上并不比一家初創公司所獲得的技術更有優勢——也就是說，網絡的分配是完全民主的，而在過去，軟件主要是由合作伙伴或軟件廠商通過最分散的銷售力量來交付。最后，產品本身也幾乎完全不同。云解決方案所面對的是一個協作、移動、開放的世界。

????某些軟硬件類別的傳統廠商或多或少被迫喪失了通過云計算為顧客提供支持的能力。比如你可以設身處地地假設自己是某家知名存儲廠商，主要向IT買家和服務提供商銷售硬件。假設現在亞馬遜（Amazon）也生產出了自己的存儲產品，并提供給現有的開發者和客戶，因為它們已經達到了相當的規模。那么你會不會決定推出一項云存儲服務，與亞馬遜進行競爭？大多數存儲領域的企業沒有這樣做，因為如此一來，他們就必須和其他客戶展開直接競爭。這樣的話，許多廠商就會處于不利地位，必須小心調整自己的業務模式，這就成了一場風險游戲。

????總之，云計算使初創公司擁有了一系列獨特的價值驅動因素，使他們可以在安全的距離內與傳統的業界豪強進行競爭。

????正如MySQL甚至沒有與甲骨文（Oracle）爭奪客戶，就兵不血刃地擊敗了后者一樣，如今的許多云解決方案也采取了相同的路數來蠶食傳統大公司的地盤。他們剛剛問世時波瀾不驚，悄然混跡于“湊合夠用”的陣營；之后，他們的產品路線圖變得越來越清晰，開始為客戶提供更重要的服務。不知不覺間，“湊合”變成了“很好”，“很好”變成了“更好”。以Salesforce.com（CRM）為例，如果關注這款銷售系統的時間超過10年，人們就會發現一個驚人的趨勢。2002年，剛剛起家的Salesforce.com只占據了全球客戶關系管理（CRM）市場2%的份額，分析人士經常把它忽略不計。他們說：“（Salesforce.com）與希柏（Siebel）和仁科（PeopleSoft）等大型CRM公司的服務完全無法相提并論。他們在功能上總是有所欠缺。”然而到了2011年，Salesforce.com已經實現了20億美元的營收入，客戶也發展到10萬多家，公司正在快速發展為世界最大的軟件公司之一。

????同樣的情節也正發生在云計算領域其他新一代的企業軟件競爭者身上。許多成本更低廉、而且從根本上更加簡單的應用已經從云計算中涌現出來，正在蠶食著傳統企業軟件公司的地盤。比如ERP云方案提供商Workday已經不再僅僅只關注中端市場，它的市場拓展得很快，最近剛與時代華納（Time Warner）和湯森路透（Thomson Reuters）敲定了合同，這必定會使甲骨文和SAP深感壓力重大。另一家云方案提供商GoodData最近剛剛募集到了1,500萬美金，因為它的2,500多家企業客戶認識到，GoodData所提供的企業情報軟件價格只相當于IBM和甲骨文的一小部分。至于我們的Box服務，客戶可以在Box上存儲內容，價格往往只相當于傳統系統的五分之一，客戶還可以在任何設備上訪問這些內容，這種功能是任何預制應用都不可能實現的。

????與此類似的是，許多新興的云計算企業提供的服務對于前一代軟件來說都是不容易實現、不可能實現，甚至是不必要的。例如Zendesk和Assistly與大量的系統實現了更輕松的整合，扮演了一個客戶接觸點的角色，因此顛覆了整個客戶支持市場。他們的服務是預制應用廠商既沒有技術、也沒有資金和手段去做的。Jive和Yammer使用了一種無形的社交和流媒體式的模式來處理傳統的電子郵件和協作系統，這當然是缺乏社交技術的微軟（Microsoft）所不具備的。Okta和Snaplogic幫助企業將他們的身份和應用數據分別進行互聯。這些服務的共通之處是什么？那就是他們所解決的問題，傳統軟件企業不僅沒法解決，甚至還沒有開始理解。還沒等傳統軟件企業回過神來，這些初創公司已經攫取了大量市場份額，而且開始向企業應用領域的腹地挺進。

????Fans of Clayton Christensen's "Innovator's Dilemma" will be unsurprised that the formula is still alive and well, even in 2011. Yes, apparently innovation still matters. To see the original effect in action, look no further than your desktop. How much desktop software do you generally use on a daily basis? Now, compare that to how many services you -- and your company -- are using in the cloud, delivered over the web. If you're like most business people, for the past two decades you've been dealing with slow, painful and cumbersome applications to help run your projects, manage your customers and share files. Installations of software tend to take months, maintenance costs more than the original product price, and unexpected delays and down time result in a massive drop in productivity. But all this changes with the cloud.

????Cloud-delivered enterprise solutions fit consistently and nicely into Christensen's framework for "disruptive innovation." They offer cheaper, simpler and often more broadly applicable alternatives to legacy models of enterprise computing. They tend to start out as low-end disruptors, bringing cost and performance advantages to over-served customers, but as these technologies mature in their reliability and sophistication, they're spreading throughout organizations and solving some of the most demanding problems. Christensen must be gloating.

????All this spells very big trouble for the traditional enterprise software leaders who plainly know they need to make the leap, but don't have the business models, DNA, or dire necessity to support the change today. Over $270 billion dollars are spent on enterprise software every year, but so far, only a fraction of those dollars is going in the pockets of cloud vendors. There just isn't much explicit incentive for the enterprise stalwarts to move that aggressively to the web. Sudden and forceful movements may confuse customers, disrupt channels, transition economics inequitably, or lead to poor product execution. Hence the predicament.

Why is cloud a disruptive innovation?

????Not all innovations are disruptive, of course. If a technology advance merely reinforces the position of incumbent vendors in a given category, it qualifies as a sustaining -- rather than disruptive -- innovation. If there are minimal changes with respect to business models and go-to-market strategies, or the technology advance is non-linear technology, existing market players are generally safe. In the memory business, for instance, improvements in performance (within a given range) are often sustainable by the leading storage providers. Whereas the introduction of an entirely new technology, such as flash storage, gives the advantage to new players who can capitalize on the opportunity and commercialize it more quickly. This, predictably, shows up time and time again.

????When the world wide web came along and Dell (DELL) launched the Dell.com website in response, it simply became another channel with which they sold the same products. The differences between a catalog and a website are insignificant if you take advantage of the Internet effectively. Yet every market is different: The Yellow Pages dropped a ball that Yelp picked up in the world's move to hypertext, not because the web alone was a disruptive innovation for accessing local information, but because user-generated reviews were a brand new way to experience local guides. And a product like Yellow Pages that already lacked a community had none to transfer over to the new world.

????Nothing in this world nothing can be said to be certain, except death, taxes, and enterprise software disruption. In the '90s, the industry went through an upheaval as the client-server model displaced the mainframe. Describing PeopleSoft, and by extension the new wave of client-server computing, the New York Times in 1995 said, "tasks are split more or less evenly between desktop 'client' computers and larger machines." This new (and now old) standard represented a cataclysmic shift in value, because applications could now be much more powerful and modern using PC standards, data could be mostly centralized, and everything would run at a fraction of the cost compared to mainframes.

????Fast forward a decade and a half, and the same large-scale change has occurred yet again with our most core business applications by bringing them to the web. Why is this such a fundamental change, and not one that the old guard can quickly latch onto?

????Well, because nearly ever dimension of the sales, marketing, distribution, and the utility of cloud-powered software is different than before. Sales and marketing are now tied to end-user adoption of services, not mandated top-down deployments of enterprise software. Distributing technology over the web offers current market leaders no intrinsic advantage that a startup can't access – that is, the web is served up completely democratically, whereas software in the past was usually delivered via partners or vendors with the most extensive salesforces. Finally, the products themselves are nearly entirely different. Cloud solutions generally embrace a world defined by collaboration, mobility, and openness.

????And in a few software and hardware categories, traditional vendors are more or less forced out of ever supporting their customers through the cloud. Put yourself in the shoes of an incumbent storage vendor that has traditionally sold its hardware to IT buyers and service providers alike. When Amazon (AMZN) builds out its own storage that it can offer to developers and customers (which it has) because of the scale they've achieved, do you decide to compete with them by launching a cloud storage service? Most in the storage space haven't, because they would find themselves competing directly with their other service customers. This leaves many vendors in the unenviable spot of having to finesse their business model like a game of Risk.

????Altogether, this gives startups a unique set of value drivers that allow them to compete with traditional powerhouses from a safe distance.

????Just as MySQL classically went up against Oracle (ORCL) without ever competing for customers, many cloud solutions today are similarly disrupting the older guard by initially slipping into the "just good enough" category. From there, product roadmaps become more elaborate, customers are served in more meaningful ways, and before you know it just good enough becomes great, and then better. Those who've watched Salesforce.com (CRM) over a ten year period have witnessed a striking example of this trend. In 2002, the then-upstart CRM had 2% marketshare of the global CRM space, and analysts often wrote it off, saying, "[Salesforce] doesn't compare well at all with other large CRM packages like Siebel and PeopleSoft… they will almost always come up short on the functionality side of the equation." In 2011, they'll hit $2 billion in revenue and 100,000 customers, putting them on track to quickly be one of the largest software companies in the world.

????The same is happening with the next batch of enterprise software players in the cloud, where lower cost and fundamentally simpler applications have emerged as aggressors to the traditional enterprise stack. Workday, no longer a mid-market-only play, has quickly shot up market, closing deals with Time Warner (TWX) and Thomson Reuters, which is sure to ruffle some feathers in Redwood Shores and Walldorf. GoodData raised $15 million on the heels of over 2,500 customers realizing they can get business intelligence software at a fraction of the cost of solutions from IBM (IBM) and Oracle. And at Box, customers can store content often for a fifth of the cost of traditional systems, and then get to that information from any device, something that is universally impractical with an on-premise application.

????Similarly, many emerging companies are operating on dimensions that were never easy, possible, or necessary in a previous generation of software. Zendesk and Assistly have disrupted the support market by integrating more easily into the myriad of systems which now act as customer touch-points, something on-premise vendors have neither the technology nor wherewithal to do. Jive and Yammer take on traditional email and collaboration systems by incorporating an untouchable social and stream-like model that anti-social technologies from Microsoft unsurprisingly lack. Okta and Snaplogic help businesses connect their identities and application data together respectively. What do all of these services have in common? They're all solving new problems that the incumbents not only can't attack, they also can't even begin to understand. But before they know it, startups will have pulled away significant market share as move more deeply into the enterprise.