Wow, how time flies and how many things have happened in the market these days to distract a market observer! Namely, only over a year after my SaaSy Discusions (Part I) and SaaSy Discussions (Part Ia) blog series, some time has at last become available for more discussions on the intriguing topic of software as a service (SaaS).
The title of this SaaSy discussions series might be somewhat deceiving, since the question might no longer be whether to go for SaaS or the on-demand computing deployment mode, but rather how to go about it for both vendors and users. Indeed, the current tough economic situation certainly has something to do with making this “go on-demand” decision a bit easier for both software providers and users.
Really, with the stock market in the US recently going even below the 6,500 mark (the lowest in decades) and unemployment heading to well over the 10 percent figure before the economy turns around (even 14 to 25 percent in some geographies and demographics), hardly anyone can disagree that it’s been tough and ugly out there. Although unemployment has been somewhat less of an issue for the IT sector (where it is likely to remain at “only” about a several percentiles level), there is the credit freeze issue. Namely, credit is nowadays either too darn expensive or non-existent, and IT departments need to reduce costs while managing their IT assets at best possible levels of performance, in whatever way is viable.
The “Do More With Less” Imperative
Many sources indicate that IT spending will be flat or down in most market segments, with a few rare exceptions in sectors such as health care, federal (but not state or local) government, bankruptcy lawyers, and perhaps some commodities, e.g., precious metals. While success is not guaranteed for anyone, including the avant-garde Web 2.0 startups, it is still important to note that even the biggest traditional independent software vendors (ISVs) are predicting a tough fiscal year (or even much longer) ahead.
One way for IT departments to “do more with less” could be to assemble more efficient enterprise systems via server consolidation, enterprise applications standardization, virtualization, or by using some proven free and open source software (FOSS) products to replace current license-based solutions (including operating systems, databases, and even the ubiquitous Microsoft Office desktop productivity components).
Pundits and on-demand advocates such as Saugatuck Technology or Amy D. Wohl also advise companies to consider cloud computing for ad hoc infrastructure, with elasticity to handle transactional volume peaks. Cloud computing is the natural evolution and extension of SaaS, and a much broader umbrella that encompasses not only the application services solution layer (i.e., SaaS offerings per se), but also all the underlying computing, data storage, enterprise application integration (EAI), security, workflow, etc., infrastructure that is also provided as a service, on demand.
One plausible idea these days is to move applications that are not mission-critical and that include data that can be successfully secured to lower-cost SaaS platforms. For instance, AMR Research recently reported that retailers are increasingly asking about SaaS offerings for product lifecycle management (PLM); price optimization; advertising, marketing, and promotion (AMP); and customer loyalty management software as they get more comfortable with the concept and look for ways to move needed functionality off capital budget lines.
The SaaS Opportunity for Users
The SaaS opportunity for enterprises starts with potential capital savings, since SaaS deployments do not require large up-front capital investment in terms of procuring hardware and software licenses. Then labor savings come from much lighter implementations (and sometimes no implementations at all) and from no need for operational expenses (contrary to a never-ending “hamster wheel” cycle of keeping track of constant patches and updates when it comes to traditional on-premise software deployments).
Finally, companies have much more flexibility to react to business and market changes, including mergers, new competitors, reorganizations, new product offerings, or new regulations in the market. Flexibility starts with the company’s ability to add and subtract users as necessary.
Furthermore, metadata-based customization (rather than via costly changes to source code, as is often the case in on-premise software) can be done either via a company’s internal IT department, by the SaaS vendor, or by its business partner. It is also much simpler for an unhappy customer to move to another SaaS product (as compared to switching on-premise providers), while some SaaS vendors also offer the option to move the SaaS product inside a firewall (on-premise), if and when necessary and preferred.
More of SaaS Opportunities
Moreover, both vendors and users can benefit from faster time-to-value. Namely, while a traditional on-premise enterprise application (or its major release) can take several months to a few years to design and implement, a SaaS application (or its next incremental release) rarely takes more than a few months. As a result, products and accompanying services can reach the market more quickly, revenue can accrue sooner, and there is a much better chance of capturing market share.
The traditional 80-20 rule (or Pareto Principle) that only 20 percent of the software features provide 80 percent of the value for end users holds true particularly for single-tenant on-premise solutions. This leaves a sizeable non-addressable market due to the prohibitive costs of providing new features per customer (i.e., to provide the remaining 20 percent of value via 80 percent of functionality).
Conversely, multi-tenant SaaS solutions lower the cost of providing new features per customer and thus extend new markets in the “long tail” area of market opportunity (in the customer distribution graph). In this tricky area of “selling lesser quantities of more items,” SaaS vendors are more likely to viably sell a large number of unique items (i.e., smaller chunks of special functionality), each to a relatively smaller number of customers (to provide some value within the “80 percent of the value” range).
SaaS vendors are also able to better monitor and track the actual usage of their offerings’ functionality, while Web 2.0 communications enable user-driven product design. Vendors are able to deliver frequent product upgrades, often without the user necessarily even being aware of it. An iterative development process works best for SaaS, and this agile development capability, bundled with more frequent customer feedback, typically results in greater customer satisfaction and higher customer retention.
Consequently, the general belief is that IT departments will move to SaaS as a cost-effective solution in a tough economy. As one example, according to International Data Group (IDG), 25 percent of new software will be delivered in a SaaS manner by 2010. In addition, almost no new software development and startup companies are being funded by the venture capital (VC) community or angel investors unless they have a compelling SaaS and Web 2.0 story.
While I personally can still think of some old fashioned enterprise resource planning (ERP) vendors that still maintain their happiness to remain largely on-premise (with maybe SaaS-ifying some parts of their portfolio), Amy Wohl claims that every software vendor she talks to either already has a SaaS application, is building (or planning to build) a SaaS application, or is looking for a SaaS company to buy.
SaaS: Facts vs. Fiction
I certainly wouldn’t want to come across here as someone suggesting that SaaS solutions will cure cancer, or solve world hunger, global warming and/or the US dependency on foreign oil problems. As I said in my 2008 blog post:
“Anyone who thinks one model will dominate for every possible use of software is just not an enterprise software connoisseur, is not a serious person, or is just an aggressive salesperson.”
The common sense is that the market will never migrate completely to on-demand, since some companies still run decades-old applications and are even happy with their mainframe computers. Some niche markets may not pay attention to SaaS for a long time, other markets cannot still figure out how to use SaaS (e.g., not enough processing power for heavy number-crunching applications), while in other sectors regulations still require the data to be kept within the firewall.
To make things even more complex, there have also been certain common assumptions about SaaS products that are based on fiction rather than on facts. Gartner recently expressed concern that some companies are deploying SaaS solutions based on false assumptions, with unnecessary disappointments and disillusionments as a result.
Gartner recently outlined the top five SaaS assumptions, starting with the misperception that SaaS is less expensive than on-premise software. The leading analyst house finds this to be true perhaps during the first two years but possibly not over five years.
As explained earlier on, SaaS works well at first because there’s no large capital investment for licenses or support infrastructure. However, in the third year (or so, depending on how much on-premise providers would charge for support & maintenance annually), an on-premises deployment can become less expensive from an accounting perspective as the capital assets used for the on-premises deployment depreciate.
The next myth is that SaaS is faster to implement than on-premises software, which is only true for simple SaaS requirements, but the IT world is becoming much more complicated (distributed and diverse). I concur with Gartner that SaaS vendors often quote time frames of 30 days to implement, but neglect to say that SaaS deployments can take seven months or longer.
As the complexity of the business process and integration increases, the gap advantage between SaaS and on-premises deployment times will narrow because a larger percentage of the deployment time will be associated with customization, configuration, and integration, which are equally difficult with both delivery models. Multi-tenant SaaS deployments are faster in that there is no need to buy and install hardware and software on premise.
But if the SaaS deployment is complex and requires several months, then these couple of weeks of saved time on installation do not mean much. However, they can mean a lot if the SaaS deployment only takes a month or two.
Part IIa of this blog series will continue with more SaaS facts and fiction, and will then start considering the major technical considerations before any vendor can embark on delivering a SaaS offering. In the meantime, your comments and feedback with regard to the opinions and assertions expressed thus far are welcome.
If you are applications users, how important are the aforementioned considerations in your software selections? For both users and vendors, what are your SaaS experiences (both in using and delivering a solution)?