Bloomberg Anywhere Remote Login Bloomberg Terminal Demo Request


Connecting decision makers to a dynamic network of information, people and ideas, Bloomberg quickly and accurately delivers business and financial information, news and insight around the world.


Financial Products

Enterprise Products


Customer Support

  • Americas

    +1 212 318 2000

  • Europe, Middle East, & Africa

    +44 20 7330 7500

  • Asia Pacific

    +65 6212 1000


Industry Products

Media Services

Follow Us

Bloomberg Customers

SaaS: The New Distress Call

For business information-technology needs, software as a service (SaaS) applications are the wave of the future. Pro or con?

Pro: Flexible and Efficient

Imagine you just bought a new shirt. It fit great when you tried it on, but shortly afterward you realize it’s scratchy.

Now imagine you take it back. The clerk says, "Sorry, no returns. But we’ll make alterations for a fee." Alterations won’t help—the entire shirt feels scratchy. But you can’t afford to buy another. So you get used to the scratchy shirt. You never like it, and you complain about it.

Not fun.

That’s essentially what happens to many companies that buy software to handle crucial processes encompassing accounting, human resources, or customer service. The software initially may be customized at great expense. Neither the business nor the vendor, however, can tell how it will "fit" in the months to come.

Sure, the vendor can tweak the software—for a fee, every time. Eventually the vendor’s attentiveness and responsiveness will wane. So the company struggles with software that can’t adapt to evolving demands.

Reality check: Business is a moving target. Needs change. Good tools are solutions, never obstacles. That’s why software as a service (SaaS) makes sense.

Enterprise software used to run only on dedicated computers. Those behemoth programs didn’t talk to each other, and they weren’t designed for change. Such constraints are, in part, why old-school enterprise software vendors developed such an aversion to changes in their products.

Meanwhile, on the Internet, technology is more creative, collaborative, and versatile than was ever possible within a closed corporate IT silo. Today’s programs not only exchange information easily but also depend on sharing and collaboration. These programs are constantly evolving and improving, offering greater flexibility for specific customization.

Even better, their shared nature offers greater economies of scale, bringing down the price. Security has vastly improved. And since it all runs over the Internet, individual end users have considerable freedom to use SaaS programs where and when they want, on devices they prefer. They can even customize interfaces and reporting.

Best of all: Companies can "try out" SaaS offerings thoroughly, at limited cost. You can be certain that the software you’ll use will fit and keep pace with changing needs.

SaaS vendors will never say "no exchanges, no returns." Their business depends not just on getting you to buy but also on keeping you satisfied.

Con: Better Options Abound

Over the past 15 years, the concept of software as a service has enabled many large organizations to simplify their operations.

Nonetheless, it is not the only, or best, way to deploy software. The computer is, after all, a universal machine that can adapt to a multitude of tasks for a wide variety of organizations. To claim that SaaS is the best way to provide instructions to a computer is to take that universal machine and restrict it to a tight, narrow role.

In managing information technology, whether that technology includes one computer or a giant network of machines, we generally attempt to minimize the cost of the technology, the amount of time required to keep our systems in working order, and the risk of disrupting our activities. All three of these factors can fall prey to the standardizing forces of SaaS.

Be they large or small, computers work on a local scale. They must meet a specific need and hence have been programmed and configured to meet that need. They can contain a special set of hardware, a unique database, and a set of programs they have to keep in balance. In spite of all the discussion about "local configuration" and "individual services," we can regularly find circumstances best served by a machine running its own set of software.

Even in its simplest forms, SaaS can disrupt local operations by forcing machines to use a standard interface when a different one would work much more effectively. SaaS could use up time for support services that would be better devoted to local needs. It can also introduce software, good or bad, that disrupts operations. I can already point to offices that have stopped using SaaS because the programs they get from their central provider either take too much time during the wrong part of the day or interact badly with special software they need for their business.

In many circumstances, organizations can benefit by conceiving of software as a centrally provided service such as that delivered by SaaS, but they should remember that the computer is a universal machine, and they should be unafraid to use such machines in unusual ways, especially in settings where locally provided software— individual programs running on specific work stations—reduces costs, shortens delays, and lowers risk.

Opinions and conclusions expressed in the BusinessWeek Debate Room do not necessarily reflect the views of BusinessWeek,, or The McGraw-Hill Companies.

Reader Comments


I have to disagree with some of the points that are underlined as being crucial SaaS disadvantages, especially, the one about not being able to integrate with internal systems.

You have to communicate the business and technology requirements up front prior to purchasing the software. If you do not have a true vision for the future and also fail to consider and communicate potential integration requirements, you are putting yourself in a tough spot.

Perform detailed do deligence and legwork ........

K. Siegel

Dear BusinessWeek Debate Room:
It is rare for me to read a position "paper" written by an academic that has "it" wrong on so many levels regarding Software-as-a-Service (SaaS).

First, Mr. Grier writes, "Over the past 15 years, the concept of software as a service has enabled many large organizations to simplify their operations." In fact, SaaS as a viable business model has been around for about seven to 10 years at best. Fifteen years ago, the Internet was used on an academic level, and it was the early to middle stages of client-server technology. Any previous attempt to offer a service-like product was in the 1980s by ADP with their service bureau. So from the outset, Mr. Grier's time line is off.

Second, Mr. Grier seems to confuse computer power and computing location with software development, deployment, and maintenance. The concept of SaaS is to deliver equal capability at a lower cost of ownership with greater simplicity and ease of use. Are there instances where an internal system makes more sense? Of course, and those instances should revolve around the core competencies of the firm and its competitive advantage. Is the firm a cost player, quality player, or first in its market? All three would suggest different focal areas of deploying technology to support their market positions. Indeed Mr. Grier suggests, "We generally attempt to minimize the cost of the technology, the amount of time required to keep our systems in working order, and the risk of disrupting our activities. All three of these factors can fall prey to the standardizing forces of SaaS." Ironically, nothing could be further from the truth. SaaS dramatically reduces TCO, simplifies processes, and can significantly improve process and activities. Are their examples where SaaS failed? No doubt there are. But I can think of (and find) many more examples where internal systems exceeded budget, projects failed to deliver value, and tangibly led to the ouster of the project sponsor or reduce earnings/eps.

Mr. Grier's facts do not support his statements which, at best, fail to deliver an accurate picture. And a startling fact remains hidden from view in his assessment of SaaS. As a service, SaaS keeps the risk on equal footing with both provider and client. If the provider fails to live up to the terms of the agreement, they can cancel the contract or, at the end of the term, not renew the agreement. By sharing the burden of risk, SaaS providers are likely to take a more strategic approach to client management, project management, and product satisfaction.

I would encourage Mr. Grier to investigate SaaS more thoroughly before suggesting it not a good way to deploy software and confusing it with computing location, computing power, and the machine (for which there is little or no relevance in the SaaS model).
Keith Siegel
Regional Manager

Steven Woods

Interesting points on both sides, and I think that it is unlikely that we'll see a world that is either all SaaS or no SaaS anytime in the next decade. The question then becomes, what areas are likely to evolve toward more (or less) SaaS adoption, and what are the dynamics that drive that? Economics is obviously key, and there are a few aspects of that. The purchasing of software on a recurring basis is one aspect of SaaS economics that has been discussed at some length. It's clearly an important driver, but I'll leave it alone for a moment and dive into another area.

One area that deserves discussion in this debate is the economics of creating and maintaining software. All software companies, whether commercial open source, or not, are in business to make a profit, and any software company that provides an offering that delivers value to customers generally attracts competitors. At some point, this leads to a normal business dynamic whereby whatever internal costs are being born end up being passed on to the customer, either through price or through cutting investment in innovation.

Internal costs in a software company are significantly driven by product complexity. A more complex product is more expensive to support, period. This complexity, however, goes well beyond the usual descriptions of complexity in software--that of having too many features and too many bells and whistles. That is definitely one aspect of product complexity, and an important one, but each operating system supported, each database, each interface all add complexity. Likewise, each historical version of the software, and each service patch that is being supported add complexity. Each hardware stack adds complexity in terms of how the application performs (or fails to perform) with certain amounts of memory, disk space, CPU, etc. This complexity requires real and significant investment in people, process, knowledge, and development. The cost of that investment is always passed on to the customer eventually, either through price, failure to innovate, or failure to remain in business.

Mr Grier makes a very solid point about highly specific systems that exist on one platform, and perform a highly specific solution. This, however, is not the norm in the majority of software systems where buyers are hoping to acquire a capability at less than the full price to build it from scratch due to the fact that the need they have is shared with other businesses.

SaaS takes advantage of this economic reality better than on premise software at a fundamental level. SaaS software maintains a single version of the software (or at least, very few versions), eliminating legacy version complexity. SaaS runs on the software stack selected and maintained by the vendor, eliminating installation/configuration complexities of multiple software stack support. SaaS runs on hardware maintained by the vendor allowing them to optimize and tune performance on one infrastructure under their full control, eliminating the complexity of maintaining a support library of hardware optimization issues.

Each of these savings manifests itself for the buyer through better prices, better innovation on the solution, or a more stable financial position of the vendor.

To be sure, these advantages will not manifest themselves in a solution that is specific to an individual business. They only manifest themselves in solutions that are relevant across a variety of businesses, such as challenges in sales, marketing, support, HR, expenses, etc. This is where we are seeing SaaS solutions gaining ground rapidly.


One of the cons that doesn't seem to be mentioned is that "confidential" data that used to be held in house is now held by a--we hope--trusted third party. To me, that's one of the major cons of SaaS. In response to Keith Siegel's remark "If the provider fails to live up to the terms of the agreement, they can cancel the contract, or at the end of the term, not renew the agreement," that's true, but have you given any thought to the cost of asking for your data back from the SaaS provider and then migrating it to a new in house system? From what I've seen, SaaS makes it very, very easy for you to migrate data to their systems--but very, very costly when you want to take your data and leave.

Arockiaraj Durairaj

Flexible and efficient is not true about SaaS. "Companies can 'try out' SaaS offerings thoroughly..." This is not possible, because this also involves time and money. Just because the SaaS vendor gives a try free of charge doesn't mean it takes no money and time. SaaS locks their customers into their technology. Thus when you leave an SaaS vendor, you leave your customization and business logic built on their platform and rebuilt on the new SaaS vendor platform. SaaS will be a big success where customization is not required. For example, e-mail and document-editing software.

Rakesh Sharma

I am not sure I understood the "Cons" at all.

SaaS as it applies to managing enterprise processes (the kind that manage HR, Sales force activities, even supply chain operations) have been steadily gaining ground--especially among the more 'cost sensitive' Small & mid-sized businesses, and even in some large enterprises.

SaaS is a mode of delivering solutions and a business model for the suppliers of these software solutions. It puts a lot of control and capabilities in the hands of the end users without the limitations of huge up-front investments.

The total risk of delivery (or at least upwards of 95%) is borne by the supplier of the SaaS solution. Not the case with the on-premise providers as is well pointed out.

It doesnt appear that the panelist from the university is very close to real-life situations, which is where SaaS solutions have been implemented - and provides huge advantages for companies that would have to otherwise cobble together their own software solutions for important business processes(since SAP & Oracle's pricepoints are beyond their reach).

Join the Debate


Participate More!

Please send us your ideas for new Debate Room topics. If you're an academic, association officer, or other industry expert and would like to write a Debate Room essay, send us a query. Questions? See the

BW Mall - Sponsored Links

Buy a link now!