Software-as-a-Service vs. ASP
The first hurdle in SaaS was to break the barrier of complexity and cost, and the undisputed leader there is Salesforce.com. They blazed the trail of simple and cheap in a hosted domain, which many consider the cornerstone of the SaaS movement. But the reality is that all the model gives us is what the ASP movement promised six years ago.
I attended the inaugural ASP Forum, and with all the hype and excitement running around back then, it’s amazing it took this long, (and a renaming, SaaS) in order to realize substantial growth of hosted applications in a subscription model. I wondered back then if the market would adopt hosted applications as a service. As I look at where things have evolved to, all we really have in SaaS-touted services are canned applications that happen to be hosted. Wouldn’t hosted components of functionality that can be orchestrated into a custom composite solution be far more advantageous? I’ll get to that in a minute.
A list of tech startups attempted to sell hosted software to companies in the late 90’s with the launch of the ASP industry, but most weren’t tuned for the Web, and also carried with them complex implementations and large costs producing a tremendous barrier to widespread market acceptance. Conventional apps hosted as services from the big boys (Microsoft, Oracle, SAP, etc.) haven’t had much success either. Again, the key being Salesforce’s attack on the complexity and cost of the traditional approach, nailing down the simple and cheap. Can the SaaS movement learn from the simple and cheap approach and evolve beyond the canned application, into the hosting of components (decomposed parts of applications), for flexible orchestration into custom, composite solutions? I think so, and that’s where SunGard is heading. Still, more on that in a minute.
With nearly a half million end-user licenses, a 63% growth last quarter, almost a half billion in revenue, Salesforce has clearly proven the SaaS model works when done right, even though it’s a canned app. With the introduction of AppExchange, it appears Salesforce understands the need to build a community ecosystem of additional functionality in the service model to continue its aggressive growth. First appearing in January, AppExchange includes more than 200 applications from partners (like Adobe, Business Objects, Google, Skype, etc…), providing applications beyond CRM, such as HR, marketing, finance and others. Clearly, Salesforce is attempting to be more of a platform for services as opposed to a canned app. This is the right direction, and where SaaS is going.
The key though is the availability of configurable components, which doesn’t appear to be the reality just yet. That’s where it needs to end up. Here’s an example of why that’s important. DuPont is a legacy user of SAP’s conventional CRM application (installed internally and custom integrated with several internal systems such as inventory and manufacturing data links), as well as a user of the SaaS CRM service of Salesforce. Since Salesforce is running remotely as a canned app there are little flexibilities DuPont is left with to make of Salesforce what it specifically needs, instead it’s limited to what the canned app does. Also, since Salesforce does not have the ability to install any software (it’s 100% hosted), there are simply some needs from some customers that can’t be accommodated. As a result, over 10,000 users rely on SAP’s system, while merely 500 use Salesforce. What if components of functionality were available in a registry and the ability to write custom components to round-out hosted services was possible? That’s where SunGard is headed.
SunGard is in a unique position as both a provider of installed software as well as a hosting entity for services in terms of it’s breadth of applications and application-components to the financial services industry. Specifically in wholesale banking, SunGard provides more comprehensive software solutions than anyone else. With such a wide variety of offerings, SunGard does not suffer from the risk of commoditization of a core offering, such as CRM in the case of Salesforce. Instead, the breadth of components can be made available as individual hosted or installed services, with the ability to be combined or orchestrated in solutions involving not only existing SunGard components but partner and even customer-defined components to ensure the unique and individual needs of each customer are realized.
This effort constitutes the next step in the SunGard CSA initiative, as it evolves far beyond an internal collaborative process and technology, and into a community of partners and customers through hosted (and locally installed where required) composite solutions. I consider two keys to the success of this CSA evolution. The first being a robust collection of components in the CSA product registry, (from not only SunGard business units but various partners as well). And, second being the ability to easily orchestrate solutions by a mixture of enablement, adding custom components and integration with existing systems. SunGard will be making these key areas available over the next 12 to 18 months in pilot scenarios, and will make general availability of this configurable SaaS concept accessible shortly thereafter.
December 18th, 2006 at 1:17 pm
Good post, and it will take a company like Sungard to make the push, though many vendors are still seeing a few webservices and a MOM interface as SOA! Breaking up applications (usually the implementation of a process) and creating services is a challenge, though creating re-usable/customizable ones is even more complex.
http://fiveclouds.blogspot.com/2006/07/components-as-service-or-is-that-jus
t.html
December 18th, 2006 at 1:18 pm
Totally agree that the name of the technology has changed and the problems it supposedly solves are the same, complexity and high costs. I am starting to believe there are cultural issues and entrenched vendor interests impeding progress in solving these problems.
December 18th, 2006 at 1:19 pm
Actually, hosting services is not the hard part, as the majority of SunGard’s revenue ‘today’ comes from hosted systems, and includes common components (collaborated), so i’m not sure i understand the point, but we’re clearly not in need of solving the hosting issue. This effort is more about decomposing our historic application boundaries into even more flexible components. Allowing more flexibility. The hosting issue is really historic at this point.
December 18th, 2006 at 1:20 pm
The comment about SFDC doesn’t really apply because it is almost guaranteed that Sungard will be offering their collaborative development solution as a single-tenant application (which in my world is as valid a software as a service model as the multi-tenant approach. Benioff, who was so right on SaaS as a concept and whose leadership really created this space - is wrong on this point).
Given Sungard’s extensive experience in providing these kinds of solutions to financial services companies, we’re not going to see the kind of scale challenges that SFDC has had to overcome.
I could be wrong, but I would be very surprised to see Sungard offer this kind of infrastructure as a multi-tenant application. Having dealt directly with the security issues involved in hosting a client’s source code in our data center, the last thing the client wants to hear is that their source is on the same system as their competitor’s.
Good luck to Sungard in this venture, I personally am glad to see another strong company enter this space.
- A competitor
December 18th, 2006 at 1:21 pm
I think the problem is more basic. The resilience of the infrastructure is still not adequate, not so long ago salesforce had outages that enraged its clients. I use hotmail as my email system outside work, and its free, but if its down well i am not paying for it, as it happens hotmail is more reliable than salesforce. Salesforce however is not cheap and getting more expensive and greedy and yet there is not guarantee over it availability.