The Origins of Cloud Computing & SaaS
This post is the first in an at least two part series on how technology changes drive changes in business models and the social fabric of those business models.
[Edit: Part 2: “Cloud Computing and Agile: BFFs”]
In this post, I will share my somewhat academic take on how the underlying drivers in information technology resulted in cloud computing and the SaaS business model. In the next post, I will demonstrate how changes in the technology delivery model (on premise/PC to cloud/SaaS) drives changes in the software company business models and company cultures.
The post below is based on a paper I developed with David Popkin and Brandon Cato at Berkeley Haas. The paper was part of a broader research program on SaaS and Customer Success.
The Origins of SaaS
Over time, the shape of the information technology (IT) industry, including its customers, major products, and structure, has been shaped by the relative benefits and costs of computing (hardware), software, and telecommunications. The enterprise information technology (EIT) segment of the industry is driven by the same benefits and costs[1]. Innovations in hardware, software, and communications alter the cost-benefit calculation for computing infrastructure, and thus EIT industry structure.
The co-founder of Intel, Gordon E. Moore, famously observed the trend for the number of transistors in an integrated circuit to double every two years[2]. “Moore’s Law” has mostly held since he first noted it in 1965, the processing-power-adjusted price of computers have fallen by more than fifteen hundred times.
The price of transmitting information electronically has also fallen. “Nielsen’s Law” states that “[u]ser’s bandwidth grows by 50% per year (10% less than Moore’s Law for computer speed)” [3]. Nielsen’s Law fits data from 1983 to 2014[4]. Through the 2000s especially, the percent of US population with fixed broadband internet subscriptions grew considerably. It first surpassed 10% in 2004 and 20% by 2006, though by 2012 the percentage grew only to 28%.[5]
Finally, software development costs have also fallen, especially with the rise of lighter-weight and interpreted languages. However, as commentators including Google’s Larry Page have noted, software is also becoming more complex and, thus, slower[6] [7].
These differences gave rise to different sets of possible computing experiences for users. Consumers and businesses have to determine the best cost-benefit tradeoff given their computing needs and options. These tradeoff decisions give rise to unique ‘phases’ in computing which each has a dominant model for delivering computing services to users.
Nielsen argued that because hardware costs tended to fall more quickly than communication costs, computer users’ experiences are limited by communications speeds.[8] This was most clearly true during the on premise computing phase. However, that in the current cloud computing phase, software development time is the limiting factor. Today, the highest value factors of production innovation focuses on improving the ability of developers to rapidly bring (good) software applications to market.
There have been four major phases in computing EIT to date: Punchcard Computing, Timeshare Computing, On Premise Computing, and Cloud Computing. In each period, businesses optimized their needs relative to computing and communication costs. As the types of costs fell asynchronously, the dominant computing model has oscillated from client-server to client-installed and back again.
Four Major Phases of Information Technology
Punchcard computing dates back to the late 1800s[9], but gained mainstream business adoption after IBM’s introduction of 80 column card in 1928[10]. In this phase of EIT, enterprise customers automated previously manual accounting computing tasks and achieved significant gains in cost, productivity, and computing accuracy.
Timesharing computing emerged in the mid-1960s. Ongoing innovations in processing speeds and communication protocols (specifically the client-server model), enabled systems in which “customers could access a mainframe computer remotely” [11]. Suddenly, even mid-sized and small companies could access computing resources on an hourly rate. As technology improved and timesharing providers because more sophisticated, business customers were able to take advantage of applications ranging from simple automating accounting computation and word processing to more complex tasks like computer-aided design (CAD) and industrial process control[12].
Personal computing was made possible by two main innovations — increasingly affordable microprocessors and the graphical user interface, which enabled less sophisticated users to productively harness the power of computing. Whereas timesharing cost $10–20 per hour, with regular users billing $300 per month, personal computing represented a much more affordable option for businesses. With the emergence of personal computers in the 1980s, enterprises captured considerable cost saving on both computing and the telecommunication cost of the client-server timesharing model.
Cloud computing leverages the continuing trend of increased processing power at ever lower costs (Moore’s Law), affordable high speed internet, and improved software architecture to enable computing resource virtualization[13]. The emergence and adoption of the internet through the 1990s and the increased access speeds of the 2000s suddenly reversed the cost trade-off which drove the shift from timesharing to personal computing.[14].
Cloud computing “shifts the location of [computing] infrastructure to the network to reduce the costs associated with the management of hard-ware and software resources”[15]. For an enterprise customer, this shifts the provision of service from a locally-installed software application to one which is provided through an internet browser and inherently networked.
While the business customer interacts only with the software provider (software as a service or SaaS), the cloud computing environment is enabled by both computing infrastructure and the software platform (PaaS) on which the SaaS application runs[16].
Infrastructure as a service (IaaS) refers to the underlying storage and compute ‘base’ layer, which may be owned by the SaaS provider or a third party. These infrastructure computers are configured with an operating system and virtualization software, so that many applications may run simultaneously upon them. For example, many applications run on computers owned, configured and maintained by cloud providers such as Amazon, Google, Microsoft, and others. In addition, IaaS resources may be used more simply to store application data and static documents (images, videos, Etc.).
Platform as a service (PaaS) provides the configured software environment to host the SaaS application and connect it with the computing resources provided by the IaaS. For example, Heroku is the most prominent of many PaaS providers which allow startup SaaS companies to rapidly prototype and iterate on applications, without construction their own platform. Heroku runs on Amazon Web Services (AWS), which serves as its IaaS provider.
While cloud computing seems to complicate the provision of computing resources and software applications, it represents a specialization across the major cost areas of information technology. IaaS providers specialize in lowering storage and processing costs and PaaS providers specialize in lowering software configuration and maintenance costs. This allows SaaS providers to focus on developing software applications that best fulfill business customer needs.
Timesharing computing was replaced by personal computing due to lower hardware costs and still high communication costs. Thus, the focus of innovation in the personal computing phase was to lower communication costs. As communication costs fell, cloud computing has emerged to take advantage of low cost hardware and low cost communication. The focus of innovation in the cloud computing phase will be the development of software architecture that fully leverages the virtual nature of the computing model and the development of specific application software that reconfigures business workflow and drives real business value[17]. Instead of simply automating basic tasks and increasing the speed computationally-heavy tasks, cloud computing will reconfigure how businesses operate and are managed.
Drivers of SaaS Adoption
It is commonly accepted that one “of the defining characteristics of internet era software is that it is delivered as a service, not as a product”[18]. Since 2006, the SaaS segment has grown more than 4X from ~$6 billion[19] to ~$25 billion annually[20]. But why has software delivery migrated from the desktop to the browser window? SaaS industry Commentators, consultants, and academics propose a number of drivers from both the perspective of software vendors and consuming enterprises.
From the software vendor perspective, the SaaS model initially allowed upstart vendors to scale quickly and support customers at a lower cost. As a 2007 McKinsey article notes, “software design and delivery models allow many more instances of an application to run at once in a common environment, so providers can now share one application cost effectively across hundreds of companies”[21]. Other benefits to software vendors include a smoothed revenue model, global distribution via the internet, and access to mid-market enterprise customers[22].
Consuming enterprises were initially drawn to the SaaS model by favorable economics including lower implementation, hosting, and communications costs[23]. As SaaS became more widespread and the sophistication of SaaS vendors improved, CIOs and other enterprise buyers have found SaaS performance superior to traditional installed software options. For example, a recent survey by The Boston Consulting Group (BCG) found “the top benefit mentioned by both CIOs and line-of-business managers was mobility: SaaS solutions are able to support multiple devices and locations”[24]. In addition, BCG also found that “CIOs…also appreciated the flexibility that SaaS solutions provide for companies that want to quickly set up and scale a new application” [25].
Thus, we can see parallels between the emergence of SaaS and Christensen’s disruptive innovation model[26] — the initial development and adoption of SaaS was driven by demand for lower costs but as it has been adopted, vendors have strengthened its competitiveness by leveraging its greater flexibility.
Today, industry analysts believe that ~20% of enterprise software is delivered as SaaS, representing a ~$25B segment of the ~$140B enterprise software industry[27]. By 2018, IDC projects that ~28% of all enterprise application software will be SaaS-based[28].
Business Model Implications of Cloud Computing and SaaS Adoption
Despite its rapid adoption, the “[a]cademic literature on SaaS is limited, with most articles focused on optimal pricing strategies under a pay-per-use model of software.” [29] However, a review of the literature does reveal a number of important proposed topic areas for both buyers and sellers.
In addition to pricing considerations[30], researchers identify a series of business model issues that software publishers must resolve as they move to a SaaS model. These include increased R&D investment[31] [32], new sales and marketing approaches[33], the importance of customer lock-in (or retention) [34] [35], and a need to move to lean (or Agile) software development process[36] [37].
While most academic literature on SaaS from the perspective of vendors has focused on pricing, most of the literature on SaaS from the perspective of consuming enterprise customers has focused on the buying process[38]. In addition to the buying process, enterprise customers must considering implementation, strategic vendor management, refocusing IT resources[39], and how to extract business value from the “improved business agility and scalability, and rapid deployment timeframes” of SaaS[40].
Up next: Agile Imperative of SaaS and the Cultural Shifts Required by Agile
What makes this fun, for me, is tracing and unpacking how the technology impacts the economics (cost), which impacts the business model, which impacts the business culture and broader social organization. In the shift from on premise to Agile, there are hundreds of rippling impacts through technology, economics, business, and society.
In the next post in this series, I’d like to examine a single thread of that wave — how cloud computing and SaaS provide the business model context in which Agile software development is clearly superior to traditional waterfall development. And because Agile is about teaming, motivation, and knowing your customers, it leads to a more open, honest, and happy workforce. In this case, technology enhances personal fulfillment. It’s not all doom and gloom. More to come…
Notes:
[1] Campbell-Kelly, Martin. “Historical reflections The rise, fall, and resurrection of software as a service.” Communications of the ACM 52.5 (2009): 28–30.
[2] Moore, Gordon E. “Cramming more components onto integrated circuits, Reprinted from Electronics, volume 38, number 8, April 19, 1965, pp. 114 ff.” Solid-State Circuits Society Newsletter, IEEE 11.5 (2006): 33–35.
[3] Nielsen, Jakob. “Nielsen’s law of internet bandwidth.” Online at http://www.useit.com/alertbox/980405.html (1998).
[4] Nielsen, Jakob. “Nielsen’s law of internet bandwidth.” Online at http://www.useit.com/alertbox/980405.html (1998).
[5] “Fixed (wired)-broadband subscriptions per 100 inhabitants 2012”, Dynamic Report, ITU ITC EYE, International Telecommunication Union. Retrieved on June 29, 2013.
[6] Niklaus Wirth (February 1995). “A Plea for Lean Software”. Computer 28 (2): pp. 64–68. doi:10.1109/2.348001.
[7] searchengineland (2009–05–27). “Sergey Brin On Breaking “Page’s Law” Of Software Sluggishness”. YouTube. Retrieved 2009–05–27
[8] Nielsen, Jakob. “Nielsen’s law of internet bandwidth.” Online at http://www.useit.com/alertbox/980405.html (1998).
[9] http://www-03.ibm.com/ibm/history/ibm100/us/en/icons/punchcard/
[10] http://www-03.ibm.com/ibm/history/ibm100/us/en/icons/punchcard/
[11] Campbell-Kelly, Martin. “Historical reflections The rise, fall, and resurrection of software as a service.” Communications of the ACM 52.5 (2009): 28–30.
[12] Bell, C. Gordon. “Fundamentals of time shared computers.” Computer Design 7.2 (1968): 44–59.
[13] Vaquero, Luis M., et al. “A break in the clouds: towards a cloud definition.”ACM SIGCOMM Computer Communication Review 39.1 (2008): 50–55.
[14] Campbell-Kelly, Martin. “Historical reflections The rise, fall, and resurrection of software as a service.” Communications of the ACM 52.5 (2009): 28–30.
[15] Vaquero, Luis M., et al. “A break in the clouds: towards a cloud definition.”ACM SIGCOMM Computer Communication Review 39.1 (2008): 50–55.
[16] Vaquero, Luis M., et al. “A break in the clouds: towards a cloud definition.”ACM SIGCOMM Computer Communication Review 39.1 (2008): 50–55.
[17] Sääksjärvi, Markku, Aki Lassila, and Henry Nordström. “Evaluating the software as a service business model: From CPU time-sharing to online innovation sharing.” IADIS international conference e-society. Qawra, Malta, 2005.
[18] O’Reilly, Tim. “What is Web 2.0: Design patterns and business models for the next generation of software.” Communications & strategies 1 (2007): 17.
[19] http://www.sandhill.com/conferences/sw2006_materials/SW2006_Industry_Report.pdf
[20] http://www.forbes.com/sites/louiscolumbus/2014/12/20/idc-predicts-saas-enterprise-applications-will-be-a-50-8b-market-by-2018/
[21] Dubey, Abhijit, and Dilip Wagle. “Delivering software as a service.” The McKinsey Quarterly 6.2007 (2007): 2007.
[22]Software and Information Industry Association. “Software as a service: strategic backgrounder.” Software and Information Industry Association (2001).
[23] Dubey, Abhijit, and Dilip Wagle. “Delivering software as a service.” The McKinsey Quarterly 6.2007 (2007): 2007.
[24] J. Pineda and J.M. Izaret, “Profiting from the Cloud: How to master Software as a Service,” BCG Perspectives June 18, 2013.
[25] J. Pineda and J.M. Izaret, “Profiting from the Cloud: How to master Software as a Service,” BCG Perspectives June 18, 2013.
[26] Christensen, Clayton. The innovator’s dilemma: when new technologies cause great firms to fail. Harvard Business Review Press, 2013.
[27] http://www.forbes.com/sites/louiscolumbus/2014/12/20/idc-predicts-saas-enterprise-applications-will-be-a-50-8b-market-by-2018/
[28] http://www.forbes.com/sites/louiscolumbus/2014/12/20/idc-predicts-saas-enterprise-applications-will-be-a-50-8b-market-by-2018/
[29] Choudhary, V. “Comparison of Software Quality Under Perpetual Licensing and Software as a Service” Journal of Management Information Systems (24:2), Fall 2007, pp 141–165.
[30] Sääksjärvi, Markku, Aki Lassila, and Henry Nordström. “Evaluating the software as a service business model: From CPU time-sharing to online innovation sharing.” IADIS international conference e-society. Qawra, Malta, 2005.
[31] Choudhary, V. “Comparison of Software Quality Under Perpetual Licensing and Software as a Service” Journal of Management Information Systems (24:2), Fall 2007, pp 141–165.
[32] Software and Information Industry Association. “Software as a service: strategic backgrounder.” Software and Information Industry Association (2001).
[33] Sääksjärvi, Markku, Aki Lassila, and Henry Nordström. “Evaluating the software as a service business model: From CPU time-sharing to online innovation sharing.” IADIS international conference e-society. Qawra, Malta, 2005.
[34] Ma, Dan. “The business model of” software-as-a-service”.” Services Computing, 2007. SCC 2007. IEEE International Conference on. IEEE, 2007.
[35] Ma, Dan. “The business model of” software-as-a-service”.” Services Computing, 2007. SCC 2007. IEEE International Conference on. IEEE, 2007.
[36] O’Reilly, Tim. “What is Web 2.0: Design patterns and business models for the next generation of software.” Communications & strategies 1 (2007): 17.
[37] Software and Information Industry Association. “Software as a service: strategic backgrounder.” Software and Information Industry Association (2001).
[38] Godse, Manish, and Shrikant Mulik. “An approach for selecting software-as-a-service (SaaS) product.” Cloud Computing, 2009. CLOUD’09. IEEE International Conference on. IEEE, 2009.
[39] Software and Information Industry Association. “Software as a service: strategic backgrounder.” Software and Information Industry Association (2001).
[40] Software and Information Industry Association. “Software as a service: strategic backgrounder.” Software and Information Industry Association (2001).
