In the last few months, I have been talking to a lot of different companies about SaaS. Initially, my discussions were primarily with ISVs , as many are considering (if not already in the process of) re-architecting their solutions for SaaS delivery; but recently, an increasing number of discussions have been with other “actors” (see diagram at the end of the post):

The primary interest of enterprises on SaaS is how to best consume it; with the very classic question: “From a consumer of the service perspective, how does it differs from SOA?” (as quickly mentioned in my
chat with Diego, at a high level it does not, even though some subtle elements subsist in some areas (more about this in a future post). But IMO more interesting is that some enterprises go further and want to embrace the SaaS model not as consumers but as providers.  Enterprises wanting to embrace the SaaS model is particularly popular with companies with a strong franchise model. The idea is to provide access to the hosted software as part of the franchise deal, allowing HQ to have more control on the software environment of their franchises (especially the one they do not own) and the franchisees to get software to run their business without large upfront IT costs.

I also talked to a new breed of companies (or actually established companies re-inventing / diversifying themselves in this new breed): the aggregators. They have two interests, one short term, one longer term.  Short term, is to act as a marketplace. Akin to what shopping malls do for retail stores, aggregators offer “marketing” services, such as a directory/catalog, a search mechanism, a unique billing interface etc. and traffic! Individual providers gain in being in a marketplace as they benefit from the “overall traffic” and have the opportunity of being part of a cross selling deal. Being part of a marketplace also helps the trust relationship required to sell SaaS. In some respect, the marketplace “brand” is passed to the individual provider.
Longer term is not only to be a marketplace but being a full fledge platform for SaaS providers. The main aspiration is to be able to compose new services from the one available. This requires a deeper “aggregation architecture” which allows the sharing of data, processes and even semantic among multiple services. There are some early attempts at this, and granted some are way more advanced than others, but to be honest, I have not seen anybody cracking to full code yet. I talked to both “pure play” aggregators, and ISVs expanding their own SaaS offering to a aggregator offering and how to successfully do this, is still a challenge.

SaaS Hosters:
Another “new breed” are the SaaS hosters. Often with origins either as ASPs of the 90’s or “classic” hosters having moved “up the stack” the SaaS hosters specialize in hosting applications delivered as a service. Instead of selling CPU, storage and bandwidth (classic hosters), they offer a shared infrastructure optimized for application hosting, with services such as billing, SLA monitoring, SLA enforcement, Billing, 1st level support etc. in other words, infrastructure than many ISVs are not interested (or don’t have the skills) in building. There are usually two topics of discussion: (a) what are the shared services that ISVs will need the most and (b) How to provide guidance to ISVs to best “plug” into the shared infrastructure.

Hosters at large (this includes: web hosters and Telcos):
These “actors” are very familiar with the concepts of OSS/BSS and many have already deployed service delivery platforms (e.g. SMS, VOIP). The challenge for many of them is to move out of the “Telco” mindset and understand what it means to have ISVs as customers. It is quite clear to me that hosters will move en masse to the SaaS hoster model described above, it is less clear which one will dominate the market. It is interesting to note that I also had a few conversations with networking equipment manufacturers and we discussed what it would mean to introduce a SaaS hosting layer as part of the equipment they sell to Telcos and Web hosters.

Service integrators and value added resellers (SIs/VARs):
Last but not least, I have had many discussions with SIs and VARs. Some are quite concerned as they see their core business (deployment and customization) being marginalized by the SaaS model. The consensus is that there are still at least 3 areas where they can add a lot of value. IMO the most valuable will be integration. Integration needs will not change (they might even increase); SIs have a lot of experience there, they will be able to have a good value proposition around integration of SaaS and “in house” (legacy?!) systems. Next will be “business consulting”, the act of deployment and customization is minimized but not the need of customization. Helping the SaaS consumers in what to customize is still a valuable service. Finally helping all the actors above “move to SaaS” should be in the arsenal of all SIs. An additional value proposition (not available to all SIs) is to offer SaaS hosting services. Several large SIs host the entire IT of very large enterprises (through multimillion dollar outsourcing deals), the knowledge and capabilities acquired in such IT outsourcing deals are applicable to a large extend to SaaS hosting.

One thing is getting very clear to me, the SaaS architecture guidance initiative we started in my team will need to increase its scope and cover more than the ISV space. Cool! J

Below is a diagram, summarizing (and slightly condensing) what I mentioned above.


saas actors diagram