JIT Case Study
HG Aerospace
"System that was developed lightening fast with excellent advice!"
View this case study
Our View of the Future
- Richard Pilton, Director of JIT Systems, November 2009
Abstract
Software as a Service (SaaS) is the software industries current big thing. By looking into the historical roots of Software as a Service, covering its evolution over the past 5 decades and assessing the current position, market penetration and opinions, we evaluate whether Software as a Service will succeed by not only being used by businesses in the short term, but, to be permanently adopted by businesses.
1. Introduction
In this paper we try to predict whether Software as a Service will continue to grow at an impressive rate, sustain a high level of uptake in commercial enterprises and therefore succeed. Application Service Provision received a similar level of hype and confidence, however never grew to the levels expected making it somewhat of a failure so we will look into why Application Service provision failed and also, look into other historical initiatives such as Timeshare Computing as a comparative service based computing offering.
In order to evaluate the success of the Software as a Service, we must be clear on what qualifies the service as a success. For the purpose of this paper, the key success indicators will be the adoption of the SaaS model into the business world, the profitability of the idea as a whole, the length of time it stays profitable and the viability of it being a permanent alternative option to Off-the-Shelf software solutions. We will compare SaaS to Application Service Provision and Timeshare Computing as they have similar attributes that will provide an understanding/insight in to the potential diversity of factors that may change its success in the future.
2. The history behind Software as a Service
2.1 Timeshare Computing
In order to get an understanding of the Computing and Software service industry as a whole, we need to look at the evolution of service computing. In approximately 1955 (Campbell-Kelly, 2008), batch data processing services became a fast growing business. The service that was initially offered required the customer to send raw data through the means of mail or messenger to the service provider who would then digitally process the data to the customer’s requirements and return the results via the same means. This service grew relatively rapidly during the 1960’s, mainly due to the advancements in communication technologies. These technologies meant that customers could have a terminal at their location which consisted of a card reader and a line printer, together allowing customers to submit jobs and obtain results. Then slightly later on, teletypes or visual display units were shipped to customers to further improve the efficiency of the experience. The turnaround time for a job request was significantly poor compared to today’s standards, but still profitably quicker in the 1960’s providing the customer with a significant competitive advantage.
Table 1: (Phister, n.d.) Cited in Campbell-Kelly (2006)
Table 1 shows the growth of the Data Processing Services (DPS) from 1955 to 1978 in Billions of Dollars. By 1969, the industry had reach 1 Billion and in the next 10 years, more than tripled to 3.15 billion. Unfortunately for the industry, their downfall turned out to be the computer manufacturers. In the early 80’s, the personal computer (PC) was released; PC’s provided Data Processing capabilities on-site without the need for the added cost of a telephone connection and could provide an almost instantaneous response. As at this point computers and applications had become much less expensive (Marshall 2001), each Personal computer would pay for itself within a year and it was a relatively maintenance free piece of office equipment compared to the mainframe computers used by the Timeshare computing service providers.
General Electric were one of the largest providers of Timeshare computing services during the 60’s and 70’s and it was arguably the most profitable computing enterprise they ever provided as a service. Lee JAN, Snively (June 2000) showed that GE were an example of a company that did not have the foresight to see that the computing industry changes quickly and violently. Not wanting to spend the estimated $1B to continue competing with IBM by manufacturing more hardware, the PC’s produced by IBM destroyed General Electric’s timesharing service and combined with selling their other computing offerings to Honeywell Information Systems (HIS), General Electric bowed out of the computing industry altogether.
It is interesting that in the history of the computing service industry there was a period where services and therefore revenues were all but killed because the businesses requiring them did NOT want outside connectivity and communication or the use of large scale infrastructures.
2.2 Application Service Providers
For the rest of the 80’s and the 90’s the factors that killed the timesharing industry reversed. Businesses wanted the use of larger IT Infrastructures so were growing there on-site capabilities. This was becoming increasingly more complex and expensive to acquire and maintain, and the communication costs were plummeting.
The reason for IT infrastructures becoming increasingly expensive was due to the colossal growth in availability and use of software that could now be installed on them. Microsoft’s Windows operating system provided users with the ability to do much more than just batch processing jobs. Word processing, spreadsheets, publishing, printing and opened a path to the idea of a paperless office. As companies wanted to get the most of their computing potential, costs were raised by the need for employed IT skills to setup and maintain not just a single computer, but a network of computers with a server.
A major contributor to the change in communication costs being the Asynchronous Digital Subscriber Line (ADSL) which provided extremely fast connectivity, only previously available by expensive leased lines, over the existing copper wire based telephone network at a fraction of the cost.
So the idea to resurrect remote computing services became an attractive option for many IT companies. However, the services had to be different this time round from the time-sharing industry. Personal computers had become very advanced over the previous 2 - 3 decades so the types of business processes they were used for were significantly different and there was, and is, a large array of software applications that everyone one is using. Therefore, instead of outsourcing specific services, the Application Service Provider (ASP) was born whose plan was to rent the use of large scale infrastructures for applications to run on and access them over the internet.
The economics of this idea meant that the customers could go back to using a relatively thin-client as they did in the 70’s which require much less maintenance and make use of the cheap connectivity offerings such as ADSL. The service provider would setup and maintain the infrastructure for the large scale applications and rent the use of them. Expensive systems provided cheaply over the internet.
In Desai’s paper (et al. 2003), they found that ASPs focused on reducing developments costs for their customers and providing access to applications that would give the customer a competitive advantage as the benefits. This turned out to be wrong as they could only offer un-customised, vanilla versions of the applications and such applications could not give any competitive advantage to their customers. It turned out that those companies didn’t just want large scale applications for them to mould their business around as this meant that all the companies in their industry would be the same; they wanted large scale applications that operated differently to their competitors, in a way that made them unique. Application Service Providers were unfortunately not software developers, therefore, although their offering was a cheap way to provide standard applications, was not the industrial requirement when put in to practice.
Erikson (August 2003) sums up the fall of the ASPs in his paper titled ‘Why Utility computing will succeed where ASPs and outsourcing failed’ with this statement:
“The ASP vendors hung their hat on a weak value proposition, namely the marginal cost savings available from hosting someone else’s applications. Users did the math and were not impressed with what they saw.”
3. What is Software as a Service?
Microsoft being one of the larger providers of Software as a Service (SaaS), define it in a paper by Gianpaolo Carraro, et al. (2006) “Software deployed as a hosted service and accessed over the Internet”
This is quite a basic description that doesn’t quite encapsulate the modern model and distinguish it from either Timeshare computing or Application service provision. Although Timeshare Computing and Application Service Provision are software provided as a service, they are not SaaS. Another, arguably more accurate description, influenced by a paper by Fontanella and Tohamy (2008) is; SaaS is a single instance of a software application and database installed and maintained on a single web-server infrastructure and supplied as a service over http through a browser to multiple customers simultaneously.
Figure 1: Different Deployment options for Enterprise Applications (Fontanella, et al. 2008)
Figure 1 is a basic diagram showing the difference between the supplies of software models. The traditional off-the-shelf method is where the application is purchased, implemented and maintained by the customer on their own infrastructure. The Application Service Provider (ASP) model is depicted by a third party company hosting the application on their existing infrastructure and allowing the customer to use it over the internet. The Software as a Service (SaaS) model shows that the application and the infrastructure are already in place, with a single instance servicing many companies. The foundation idea of this is to enable the customers to benefit from the economies of scale, i.e. the more companies using the application, the lower the cost of the infrastructure per user and therefore presumably the service.
The SaaS and ASP model are often confused, but, another difference between the two is that the SaaS vendor is also the developer of the application. Application Service Providers were typically Internet Service Providers that would purchase a license (or use the customer’s license) for the desired software and install it on an infrastructure they have assembled demonstrating how, in hindsight, they were really just agents and not capable of genuinely benefitting from the economies of scale as there still needed to be an installed instance of the software per customer. Admittedly, the ASP’s could have many instances of the same software on the same infrastructure, but the main point is that the applications were never developed to have many customers/companies using the same instance, whereas SaaS applications are.
3.1 Utility computing
Utility computing was one of the first terms used in around 2003 (Eriksen, 2003), for the model that we now know as SaaS. The naming was based around the industry’s ideal that computing should and will be purchased as a utility like Electricity or Gas and regarded at the same level of necessity. This opinion of SaaS being treated like a utility is still held by key figures in the IT industry as demonstrated by a recently documented public discussion between Werner Vogels: CTO Amazon, Greg Olsen: CTO and founder of Coghead, Lew Tucker: CTO of Cloud Computing Sun Microsystems, Greg Badros: Senior Engineer Google, Geir Ramleth: CIO Bechtel, Steve Bourne: CTO El Dorado Ventures (Creeger, June 2009).
Even though the term was used in 2003 to describe the latest model of Software as a service, the idea of using computing services like a utility was the vision Leonard Kleinrock (Buyya et al. 2009) who was one of the chief scientists of the original Advanced Research Projects Agency Network (ARPANET) project which seeded the internet, in 1969 he said “As of now, computer networks are still in their infancy, but as they grow up and become sophisticated, we will probably see the spread of ‘computer utilities’ which, like present electric and telephone utilities, will service individual homes and offices across the country”
Can SaaS really be compared to utilities like Electricity and water? Surely those utilities are very simple with a fixed requirement and a fixed method of delivery whereas Software applications are complicated with the requirements being different for every customer. Electricity is not really tailored to supply a specific flavour of electricity for a specific purpose or process, it’s just available or not, on or off. However, just because software is more complex, does not mean it cannot be a service.
4. Discussion points for the success of SaaS
4.1 Is SaaS cheaper?
Considering the commonly accepted argument that businesses hold the implementation costs and timeframes as a significant influencing factor when deciding on software solutions, we can safely assume that reducing the cost of these factors is indicative to whether SaaS is actually cheaper. Combined this with an assumption by Fan, et al. (2009) where they cited PR Newswire (2006) performing a study of over 600 companies founding that SaaS did indeed significantly reduce the implementation cost and time. Therefore, this is evidence to suggest that SaaS is in fact cheaper. This could be attributed in part to the recent and current success rate of SaaS’ penetration into the business market.
Most sources agree with this perspective, however, not all. Marks (2008) argues that in the long run the SaaS model is not cheap, there may be some point to his arguments about the ongoing costs that SaaS enforces compared to a one off cost.
Microsoft at least agree (Gianpaolo Carraro, et al. 2006) that the financial side of SaaS does easily reduce the Implementation costs, however, when the ongoing costs of SaaS is applied to large lengths of time and the potential for companies to grow, the economies of scale are less clean cut. Considering Microsoft is one of the larger companies entering in to the world of SaaS, admitting this is counterproductive for the business which implies a sense of credibility to the statement.
Combining these two points, it is difficult to decisively announce that SaaS is cheaper, or indeed more expensive. In the same way that a hire purchase on a car is definitely more expensive in the long run, but much more affordable to benefit from immediately. This may suggest a reason for SaaS being mainly used by small to medium enterprises at present as they may not have the ability to purchase a software system outright.
4.2 Is SaaS better quality?
To ensure that we are making a good comparison of quality between traditional off-the-shelf software (OTS) and Software as a Service, rather than focusing on a static ‘point in time’ comparison of a selection of applications covering both, we need to look at the potential and scale of quality improvement increments applied by each scheme. This is because all software applications are not static in their offerings; they evolve over time and respond to their deployed environments and users. Therefore, the age of any software application can heavily affect the quality and quantity of specific functions and SaaS being the younger initiative as yet does not have such a diverse portfolio.
Fan, et al. (2009) found during their paper evaluating short-term and long-term competition between traditional, off-the-shelf software (OTS) and SaaS, that SaaS has the potential to be of larger increments of quality mainly due to there being a single managed instance of the software maintained by the software provider itself. However, it is also mentioned that this is based around two factors which could be considered faults to the theory. 1) The rate of defection from OTS stays at a relatively constant rate and 2) the technology used to develop the SaaS application remains the focus of innovation in the IT world.
HTML, CSS and AJAX are the main technologies used to deliver SaaS and currently do not have the flexibility of desktop software, but then again, it is not structured in the same way and it’s method of delivering its functionality is far superior when it comes to evaluating the pre-requisites in terms of hardware, software and skills. For example, a SaaS application requires a modern web-browser and an OTS application requires at least a server and a relational database management system.
In a paper comparing quality of Software as a Service and OTS/Perpetual Licensing applications, Choudhary (2007) found that by analysing how new software features are disseminated in the two software delivery methods, the SaaS model earns larger profits in the long run, therefore, assuming that they re-invest the same proportion of the income into the quality increments, should provide larger leaps in quality improvements.
Another factor that is mentioned in the same paper, due to the infrastructure of the SaaS model, the vendor does not have to accommodate such a large array of deployment environments (i.e. operating systems of the client computer) or focus a lot of their time on optimising the installation experience, as these are not recurring issues when the software is provided as a web based service. Similarly, new versions and enhancement patches do not need to managed by the client and supported by the supplier in SaaS as the customers are all using the same instance and managed by the supplier. So these factors suggest that, theoretically, there is a larger percentage of the profit that can be invested into quality increments.
4.3 Is SaaS Reliable?
A sub-set of quality is the reliability of an application. For remote software to maintain a good name in the industry, the service providers need to aim for a negligible level of down-time. Although internet connectivity continues to get better and more reliable as time goes on, it is still not 100% reliable, therefore, a company using SaaS as part of their key business processes can find themselves disabled if their internet connection goes down and this type of outage does not fall inside the Service Level Agreement of the SaaS vendor.
It’s not only the internet connection that can affect down-time, in February 2009, Google’s web mail service went off-line for approximately 3 hours (Clark 2009). This was allegedly due to a miss-communication internally which resulted in the wrong servers being taken down for maintenance. If their service was an ERP package for a company, 3-hours outage may have had severe financial repercussions per customer. Arguably, OTS applications are not immune to the same outages as companies with multiple locations often store their data on a privately owned central location which relies on an internet connection and OTS applications sit on computers that also require maintenance. This, however, does not cancel out the problem entirely as businesses are often more annoyed with problems that are not their own fault.
5. What is the current Success of SaaS?
The SaaS market is described by Bert Latamore in his article titled ‘Nine things you need to know about SaaS’ (October 2007) and agreed by David Linthicum’s article ‘How Mature is tha SaaS Market’,
“The market is in its early high-growth phase, having passed the inflection point in the typical high-tech market scenario,...”
The idea of the “High-growth” phase is supported by the financial statistics displayed in figure 2 and figure 3 which shows the Year over Year (YoY) gain in the second quarter of 2009 for both SaaS vendors and OnPremise Vendors (OTS). This compares the current pattern of growth with respect to the largest suppliers of the two different software delivery models. Despite the current economic issues the world is experiencing, the SaaS model is experiencing positive YoY gains whereas the OTS suppliers are feeling the downturn.
Figure 2: Software Insider Index Q2 CY 2009 SaaS Vendors
Figure 3: Software Insider Index Q2 CY 2009 On Premise Vendors
Revenues announced by AccountingWeb (May 2009) showed that the overall SaaS market had reached revenue of $6.6 Billion in 2008 and projected it to reach $8 Billion in 2009, which is consistent with the YoY gain figures above. Estimates by many sources have been between $11.2 Billion and $13.9 Billion in 2012, however, these estimates are only based on the recent success of SaaS and not reliable as an ongoing trend, they are not able to account for unforeseen circumstances or other industrial changes that could occur.
The next comment by Bert Latamore (October 2007) on the SaaS market is;
“...It's characterized by large numbers of fairly small vendors, with more entering constantly. In this case, the growth in the number of providers is being aided by some very large organizations, including Microsoft Corp. and IBM, and some small middleware vendors such as Progress Software Corp., which are helping business partners, particularly independent software vendors, move into the market”
The backing by large organisations in a specific initiative can often be mistaken as a certainty for success as in 2000 (Eriksen, 2003) the ASP model was heavily endorsed and supported in the same way by AT&T and Sprint (among others) both being giants in the communication sector.
One of the most famous success stories of SaaS is Salesforce.com who provides web-based CRM solutions. Salesforce.com started in 1999 (www.salesforce.com, Sep 2009) and grew steadily until around 2006 when their financial growth rate took a significant increase. In 2009 Salesforce.com has increased its total revenues by 50% each year for the past 3 years. The figures filed on February 2009 showed total revenues of nearly $1.1B (www.businessweek.com, Sep 2009) and is expected to increase another 50% during the coming year. As it stands, Salesforce.com services in excess of 63,000 businesses (www.salesforce.com, Sep 2009) across the world and the larger companies (e.g. Fujitsu) are beginning to join the trend.
Figure 4: Screenshot of the SalesForce web application
6. What is the future of SaaS?
Considering the current activities of large, well-known software vendors, it would be easy to declare that all software applications that are currently used will be available through a web based offering. SAP are releasing a SaaS version of their ERP system (www.sap.com, 2009), Sage has released a web based version of their CRM/ERP application, Microsoft have Office Live and many of the larger web-design companies are now trying to provide their Content Management Systems as-a-service, which incidentally is one of the fastest growing services behind CRM. But, this would almost certainly be simply a re-invention of the ASP using the benefit of hindsight and learning from previous mistakes, but SaaS is not ASP version 2.0, there is a much larger plan behind SaaS, it is but the tip of the ice-berg.
The future of SaaS is Cloud computing. In July 2008, a joint venture between Hewlett-Packard, Intel and Yahoo announced their plan for the future of IT (Furguson, 2008). Heavily influenced by the success of the SaaS model, the aim is to provide more than Software as a Service, also Platform as a Service (PaaS), Infrastructure as a service (IaaS) and Human as a Service (HuaaS) grouped together and called by some; Everything as a Service (XaaS).
The SaaS market has been taken up predominantly by small and medium sized enterprises as for smaller numbers of users means less monthly cost making the return on investment all the more clear. But not only this, the smaller enterprises don’t have the infrastructure in the first place to be able to use applications of this scale or the finances to purchase and implement the applications, where as the large companies do. With providing the platform, the infrastructure as well as the software over the internet as a service, the fundamental make-up of how we use computers will change. It is not the aim for companies to be reliant on local in-house IT expertise, a local in-house infrastructure or indeed the capabilities of the personal computer on everyone’s desk, but to go back to something not dissimilar to the terminal computing that was in use in the 70’s. Our PC’s will become the thin-client gateway to the capabilities of the ‘international computing network’ currently known as the internet soon NOT to be known by some as the ‘information super highway’ because providing information is a small sub-set of its purpose.
The following definitions are provided by Lenk et al. (2009) paper on providing an architectural map of what is cloud computing.
IaaS is described with 3 types of Basic Infrastructure Services (BIS): Computational, Storage and Network. A good example of one of these is GoogleFS which is a distributed file system developed by Goolge Inc. demonstrating the beginning of the reduction in hard disk space importance for local machines
PaaS can be classified into 2 different types of services; Programming Environments and Execution Environments. Both Microsoft and Sun Microsystems have programming tools and runtime environments to sit on top of IaaS.
HuaaS is the newest addition to the cloud to encapsulate those services that require human intelligence interaction. For example, newsworthy broadcasts, political predictions or decision based aggregation of information/opinions.
These are merely the building blocks to achieve the vision. Buyya et. al. (2009), describe the potential of industry specific clouds which are an accumulation of many service providers which strategically complement each other. Therefore, depending on the sector or industry a company trades in, there will be a cloud existing where many relevant web-based applications will be available, all of which can interact with each other on demand at the customer’s request. This will evolve into global Cloud exchanges or markets where for example the retail cloud will be able to interact with the procurement and shipping cloud to arrange collection and deliveries. Accounting functions will become all but fully automatic. The possibilities are somewhat endless.
Other inherited by-product benefits will emerge, for example, loyalty card schemes currently in wide use by every major retail chain in the world hold masses of detailed information about purchasing trends where as with Cloud computing, global usage statistics will be available in fine detail across all commercial and consumer activities and potentially available to anyone paying for the service.
All of this sounds somewhat amazing, however, this is far in the future assuming that the vision is maintained and shared by a large percentage of the world. Returning back to the more immediate future, shown in figure 5, Google has heavily embraced all levels of the cloud computing phenomenon and provides a long list of SaaS, Paas and IaaS offerings. They have partnered with IBM (Baker, 2007) to compete with the other IT giants in a move to ensure they are not over-taken.
Figure 5. Lenk et al. (2009) the make-up of cloud computing
One of Google’s cloud computing applications is Google Docs. Google Docs is a good example of where Cloud Computing presently stands, in terms of maturity, compared to the applications we are used to using. It crosses pretty much all levels of the cloud computing as it is a software service standing on its own wed based platform enabling the user to store files in its web based infrastructure. The Google Docs service, although easy to use and attractive in its idea, is only capable of manipulating documents and spreadsheets to the standard offered around 15 years ago by applications like Microsoft Office, therefore, in today’s commercial reality, only serves as a discussion point of what is to come.
Figure 6: Source: docs.google.com (2009)
7. Conclusion and future discussion
A constant issue throughout all the perspectives looked at here with regards to Software as a Service is that large companies are not as quick on the uptake of the model, this needs to change in order for SaaS to convert to a genuine qualified and trusted option. This “large company” problem seems to be predominantly due to a combination of the cost implications and the flexibility to provide a competitive advantage.
With respect to the cost, the maths do not appear to be as clean cut when a company is already large enough to achieve large scale systems themselves without the need to pay ongoing subscription charges. Especially as the current subscription cost model is based on a per seat payment structure and larger companies have a lot of seats. Usually in the IT world, the larger companies are the first to take on new initiatives whereas SaaS is growing from SME’s upwards. This does not necessarily need to be a problem - once the applications supplied through SaaS are mature enough in functionality to compete with their OTS rivals, the SaaS vendors will need to re-model the charging structure specifically for large enterprises, so that it is more of a one off cost option in order to make them more competitive with the OTS vendors. SaaS should have the advantage by actually being cheaper considering they will already have the additional revenue of a huge SME income and not need to rely so heavily on the larger customers.
With respect to the flexibility of SaaS applications being able to provide competitive advantage to each individual customer, this is certainly a big issue, but not one that can’t be overcome. This is one of the major reasons why Application Service Provision failed. SaaS is often viewed as a much less flexible software solution to OTS as there is only a single instance of it, therefore, by using it, companies may be conforming their business processes to a standard, and effectively losing their individuality. SaleForce.com has launched their new product called “Force.com Platform” which enables customisation of its SaleForce application which does appear to be at least part of the answer. The ability to perform extensive customisation is certainly going to be an absolute requirement for SaaS to succeed.
However, even if it becomes a standard that SaaS vendors develop the ability for companies to customise their experience, who is going to do it? As the initiative is in its infancy, there is a relative shortage of consultants who are experts in the field of SaaS which will need to change quite rapidly. Consultants and practitioners are required to path the way of using SaaS combined with bespoke service oriented architecture in order to reap the benefits of cloud computing while still maintaining whatever competitive advantage through individuality. As there are many consultants for applications like SAP and Sage, upon their release of Service based versions of their applications, the availability of SaaS experts as consultants may take a significant up-turn.
So, as long as SaaS vendors provide extensive customisation capabilities that breed a generation of SaaS consultants and conform to a shared global standard, eventually, even large companies will be able to focus on customising the functionality of a SaaS application without the expense of the infrastructure and the maintenance. Considering these factors, in answer to the question “Will SaaS succeed?”, the answer is yes because the current profitability of the model is at a level far greater than that of the ASP, at a level which will enable it to mature enough to resolve the issues of cost and flexibility for companies of all sizes.
As for SaaS continuing to succeed, the future vision of Saas becoming true cloud computing seems to also rely on two governing factors:
1. All commercial enterprises agreeing that it is the way forward and all SaaS vendors working in the same direction with common standards of development creating common communication channels.
2. There are no changes in the future which affect the foundation principles on which the Cloud computing idea is built upon.
With respect to factor no. 1. Even though all the large IT companies (Google, Microsoft, Sun Microsystems, etc) are ploughing lots of research, development and funding into cloud computing, this still does not mean for sure that it is ongoing success. Large companies who are not in the IT industry have often displayed a dislike to conforming to centralised schemes or markets as it allows smaller companies to benefit from their success more than benefitting more from it themselves. For example, comparison websites are fantastic for smaller insurance companies that want to provide the cheapest deal to as large a share of the market as possible, the comparison website provides the route to market and the insurance company needs only focus on cheapness of their offering, but the large insurance companies who already have a large share of the market refuse to compete, and advertise the fact that they do not participate, then counteract the initiative by offering subtly different and more enticing deals through their existing channels. Therefore, there is potential for the Cloud computing vision to be no more than an idea.
However, for the same reason that SaaS will succeed in the short term, will enable it to continue to succeed. The momentum of the existing profitability is a considerably large factor. There has been a lot of evidence to suggest that the SaaS model allows the vendors to re-invest a larger percentage of their turnover for further development allowing them to mature quicker. Also, the one unspoken benefit SaaS is the integration potential. When there is only one instance of an application for many companies, one integration project integrates many companies. As soon as there are many SaaS applications offering the ability to inter-communicate, there will be an exponential increase in its attractiveness. For example, if any SaaS based CRM applications could seamlessly integrate with the any SaaS based ERP system and with any SaaS based accounting system to provide and retrieve information on demand and all customers also using those applications are able to send and retrieve invoices, delivery notes, quotations, payments and purchase orders automatically from their own system, all companies will instantly inherit business to business integration. This is something that currently is only available to some large companies who have invested considerable amounts. The cost savings will simply become irresistible, even to the largest of enterprises.
With respect to factor no. 2. In order to evaluate whether or not there could be any future changes that will affect the foundation principles of Cloud computing, we can only look at the Timeshare computing industry during the 60’s ad 70’s to see if we can predict any issues that may arise. Its downfall was caused by a significant reduction in the cost of computer hardware and a high cost of telecommunications. We could confidently say that communication costs are low and will stay low, however, that may be a mistake. Our internet connections are currently of insignificant costs for the amount of data we use, if the traffic and the speed requirements significantly increase as a result of SaaS, the ADSL and SDSL solutions may prove to be insufficient which would leave only a leased line option. A leased line with speeds far greater than DSL technologies is a considerable investment often costing in the region of 10’s of thousands a year and 10’s of thousands to install, especially if an office is in a remote location, therefore, the Cloud computing initiative may suddenly become unrealistic for the smaller companies which is, as we have seen, the main source of its success.
The cost of hardware is a pretty stable market and will probably not affect Service computing in the same way that it did in the 70’s, however, the more modern issue is the power requirements and the carbon footprint. Google for example could potentially increase its power output if there is a large uptake of its storage, database and application services. We can only guess the increase in power requirements as many times more than it currently uses, but as of right now their data centres consume as much power as the city of San Francisco (Buyya et al., 2009) which has a population of 7.5 million. If service applications cannot maintain appropriately efficient usage of power, a potential future problem may become more political than commercial.
Will SaaS maintain its success and continue to be a serious alternative to OTS applications? I believe this is heavily dependent upon SaaS applications maturing to a level that will enable them to provide the same functional capabilities that all other software can. They cannot survive on advertising cost savings alone as the functionality gap will always be a problem and larger companies prefer superior functionality. Superior functionality provides a competitive advantage and therefore the choice is between the ability to earn more money through OTS at a higher cost or less money at a lower cost through SaaS.
I don’t believe there will be significant enough changes to the foundation principles to make the idea fail. If Cloud computing grows to a scale where the internet connections are not good enough or the power consumption is too great, other industrial players will ensure that the problem is solved once they realise the sheer magnitude of profit that could be made.
References
AccountingWeb (May 2009), ‘SaaS revenue projections, mega trends in 2009’, Available at: http://www.accountingweb.com/item/107700
Baker, Stephen, (December 2007), ‘Google and the Wisdom of Clouds’, Business Week, December issue Available at: http://www.businessweek.com/magazine/content/07_52/b4064048925836.htm?chan=magazine+channel_top+stories
Buyya, Rajkumar., Yeo, Chee Shin., Venugopal Srikumar., Broberg, James., Brandic, Ivona. (2009), ‘’ Cloud computing and emerging IT platforms: Vision, hype, and realtiy for delivering computing as the 5th utility, Future Generation Computer Systems, Vol. 25, P. 599 - 616
Campbell-Kelly, Martin., Garcia-Swartz (Jan-March 2008) ‘Economic Perspectives on the History of the Computer Timesharing Industry’, Annals of the History of Computing, IEEE, Vol. 30, Issue 1, P. 16 – 36
Campbell-Kelly, Martin. (May 2009) ‘The Rise, Fall, and Resurrection of Software as a Service’, Communications of the ACM, Vol. 52, No. 5, P. 30
Carraro, Gianpaolo., Chong, Fred., (October 2006) ‘Software as a Service (SaaS): An Enterprise Perspective’, Building Distributed Applications, msdn.microsoft.com, Article aa905332
Choudry, Vidyanand. (Fall 2007) ‘Comparison of Software Licensing and Software as a Service’, Journal of Management Information systems, Vol. 24, No. 2, P. 141-165
Clark, Lindsay., (Apr 2009) ‘Cloud Computing: cloud is good news for small companies’, ComputerWeekly.com, Available at: http://www.computerweekly.com/Articles/2009/04/14/235622/cloud-computing-cloud-is-good-news-for-small-companies.htm
Creeger, Mache. (June 2009) ‘CTO Roundtable: Cloud Computing’, Distributed computing, Vol. 7, No. 5
Currie, Wendy L., Joyce, Philip., Winch, Graham., (2007), ‘Evaluating Application Service Provisioning Using System Dynamics Methodology’, British Journal of Management, Vol. 18, P. 172-191
Desai, Bhavini., Currie, Wendy., (2003), ‘Application Service Providers: A model in Evolution’, ACM International Conference Proceeding Series, Vol. 50, P. 174-180
http://docs.google.com/ (Sep 2009)
Duval, M. (May 2000), ‘ASPs: the good, the bad, the ugly’, Interactive Week, http://www.zdnet.com/intweek/stories/news/
Eriksen, Leif. (Aug 2003) ‘Why Utility Computing Will Succeed Where ASPs and Outsourcing Failed’, www.utilitycomputing.com
Fan, Ming., Kumar, Subodha., Whinston, Andrew B., (April 2008), ‘Short-term and long-term competition between providers of shrink-wrap software and software as a service’, European Journal of Operational Research, Vol. 196, P. 661-671
Ferguson, Scott., (August 2009) ‘The future of cloud computing’, eWeek, P. 24
Fontanella, John., Tahomy, Noha., (March 2008), ‘What You Need to Know About SaaS’, Supply Chain Management Review, Vol. 12, Issue. 2, P. 10-11
Kakabadse, Andrew., Kakabadse, Nada., (2002), ‘Application Service Providers (ASPs): New Impetus for Transformational Change’, Knowledge and Process Management, Vol. 9, No. 4, P. 205 - 218
Kontzer, Tony., (June 2009), ‘the Cloud’s the Limit With Anything as a Service’, Baseline Mag, P. 24-27
Latamore, Bert, (October 2007), ‘Nine things you need to know about SaaS’, www.computerworld.com, Available at: http://www.computerworld.com/s/article/9042620/Nine_things_you_need_to_know_about_SaaS
Lee JAN, Snively (June 2000), ‘The Rise and Fall of the General Electric Corporation Computer Department’, IEEE Annuals of the history of computing, Vol. 22 Issue. 2, P. 53-60
Lenk, Alexander., Klems, Markus., Nimis, Jens., Tai, Stefan., Sandholm, Thomas., (May 2009), ‘What’s Inside the Cloud? An Architectural Map of the Cloud Landscape’, Proceedings of the 2009 ICSE Workshop on Software Engineering Challenges of Cloud Computing, P. 23 – 31
Linthicum, David, (October 2007), ‘how Mature is the SaaS Market?’, Intelligent Enterprise – Information Week, available at: http://intelligent-enterprise.informationweek.com/blog/archives/2007/10/how_mature_is_t.html
Marks, Gene. (July 2008), ‘Beware the Hype for Software as a Service’, BusinessWeek Online, Business Source Premier, P9
Marshall, GE., (Feb 2001), ‘Back to the future?’, Agency Sales, No. 2, P. 37-41
Patnayakuni, Ravi., Seth, Nainika., (2001), ‘Why License When You Can Rent? Risks and Rewards of the Application Service Provider Model’, Special Interest Group on Computer Personnel Research Annual Conference, M. Serva, Ed. SIGCPR '01. ACM, New York, NY, P. 182-188
Salesforce.com (2009), ‘CRM Company Information’, Available at: http://www.salesforce.com/uk/company/
Reuters, (2009), ‘Salesforce.com’, available at http://investing.businessweek.com/research/stocks/financials/financials.asp?ric=CRM
Wang, Ray, (September 2009), ‘Quarterly Financial Tracker: Q2 CY 2009 SaaS Vendors and Purpose Built Solutions Rack Up Big Wins’, A Software Insider’s Point of View, Available at: http://blog.softwareinsider.org/2009/09/28/quarterly-financial-tracker-q2-cy-2009-saas-vendors-and-purpose-built-solutions-succeed/
www.sap.com, 2009, ‘Get started with SAP business BYDESIGN’, available at: http://www.sap.com/sme/howtobuy/businessbydesign/index.epx