It is always amusing (and illuminating) when different events converge about a common issue or problem.
I the last few days, I have read an article demanding that multitenancy is the only way to go for SaaS applications and that ISVs with existing applications must make the substantial investments required to change their traditional on-site application into a vibrant, cost-effective SaaS application.
I guess I'd say I semi-agree.
- Most experts in SaaS and Cloud Computing agree that a multi-tenant application is better than running multiple instances of single-tenant applications.
- Almost all new SaaS applications (note almost, not all) are built to the multi-tenant model. If an ISV is going to compete with them, he should be architecting and building to the multi-tenant model, too.
- But (you knew that was coming), some ISVs are providing very specialized, mature, robust applications to niche markets. If their goal is simply to provide another software distribution option for existing and new customers, they may find that the investment in multi-tenancy is not justified. Especially if they're in the process of bringing out a substantially enhanced or entirely different product and they can wait for multi-tenancy until the next generation or next new product. And yes, ISVs wth existing products who want their on-site and on-line products to be compatible face other substantial development chores if they want to move to multi-tenancy. Again, who your customer is, whether there is competition (or whether there might be competition soon), and what your resources and goals are, all count.
And then, in the very same week, I saw a really interesting product that attempts to answer the traditional ISV's need to provide a true, multi-tenant SaaS product, without the need to do the extensive development investment. SaaSGrid from Appendra offers ISVs a SaaS environment (complete with infrastructure, metering, billing, etc.) where your existing application can be programatically dissected and reappear as a multi-tenant SaaS application.
This is a very clever and timely idea. I spoke with Sinclair Schuller, Appendra's CEO, and he told how he and some of Appendra's other team members had built several earlier SaaS programs, spending most of their time in building the common elements every SaaS program needs, and only a much smaller portion of their time focused on the application itself. It occurred to them that it might be better to build that infrastructure as a business offering and let application vendors simply plug it in. Sinclair pointed out that when he was building those applications if there had been something to buy that would have provided this substantial development shortcut (and significant cut in ongoing support and management costs), he would have been happy to use it.
Appendra has already attracted both ISVs looking to move their commercial applications to the cloud and enterprise customers looking to turn some of their legacy applications into cloud-based internal offerings that can be delivered and supported for geographically dispersed user populations.
I suspect that ISV's desire to move their offerings to the cloud, combined with a better understanding of just what a SaaS applications involves, will make offerings like Appendra very attractive. As far as I now they are a unique offering. There are lots of vendors offering ISV partners assistance in moving to the cloud, but this particular combination of offerings is the first I've seen. Of course, I'm assuming I'll see more soon.