The Boomi Blog

Multi-Tenant Integration – Busting the Myth of “Hosted” versus SaaS

I’ve noticed a trend recently of conventional integration vendors offering a “hosted” alternative to their on prem products and appliances. Unfortunately, those offerings are then marketed as “SaaS” or “on demand” simply because they are hosted “in the cloud.” To be clear, that is an ASP model – a model that did not fare so well. Just because something is hosted does not make it SaaS. (Nor does leasing or offering monthly payments make it SaaS either.)

So what’s the difference and what’s the big deal? The chief characteristic of SaaS that separates it from the old ASP model is multi-tenancy. Multi-tenancy simply means there is only one copy of an application deployed in the cloud (single instance) but that all customers can use that copy (multi-tenant) and even customize it to meet their unique requirements. Multi-tenancy is what makes all the great values of SaaS, such as rapid time to value, faster innovation cycles and ultra-low cost structure, possible.

It is particularly important in the world of integration because integration is such a fundamental building block for the SaaS community. Think of any common infrastructure system such as electricity as an analogy. It only makes sense to build common infrastructure systems once to the benefit of the entire community.

In his book “The Big Switch,” Nicholas Carr describes how one hundred years ago, companies stopped generating their own power with “dynamos” and plugged into a growing national power grid of electricity. Looking back today, the benefits are obvious: dramatically lower cost, greatly reduced maintenance, and ubiquitous distribution. It also made the process of upgrading technology much easier as changes made to the common grid were immediately available to the benefit of all users. But most importantly it addressed the scalability issue that was created by the limited reach of isolated dynamos and in the process unleashed the full potential of the industrial revolution.

We are in the midst of a similar revolution today with the advent of SaaS and cloud computing and integration has become the modern-day version of electrical power – call it digital electricity. Conventional integration products and appliances are single tenant – they were designed and built to support individual enterprises much like the electrical “dynamos” of yesteryear. In the single-tenant model, the burden and cost of building, implementing, and maintaining integration is shifted to the end customer. Every business must generate its own digital electricity at great expense. And attempting to grow the multi-tenant SaaS ecosystem with single-tenant integration products will result in sky-rocketing maintenance costs and greatly limited adaptability and scalability.

When we built Boomi On Demand, we recognized that the industry needed a “universal power grid” to unleash the full potential of the SaaS revolution. We built the Boomi platform from the ground up as a true, single-instance multi-tenant SaaS platform. It was designed and built to natively handle the complexity of multi-tenant SaaS applications. ISVs and end customers alike can now plug directly into Boomi’s “power grid” and have instant access to integrate with the industry’s largest network of interconnected SaaS, Paas, on-premise and cloud compute environments.

This is not a perfect analogy but I think it makes the point. Multi-tenant platforms such as Boomi unite and power the growth of the SaaS ecosystem. Single-tenant integration products simply weren’t designed with SaaS in mind – they were built for conventional on-premise implementations.

So yes, multi-tenancy does matter. We have the opportunity as we build out the SaaS ecosystem to avoid the integration mistakes that plagued the enterprise era. Adopting a common, multi-tenant architecture for integration is imperative to support rapid growth and expansion.

SaaS Apps and Conventional Integration Products - Oil and Water

Most analysts covering the SaaS industry today would agree that integration is the number one barrier to the adoption of SaaS technology. The problem lies not in SaaS technology itself but in the attempt to use conventional integration products and appliances for SaaS integration. It’s a technology mismatch. Conventional integration products were built for traditional on premise software implementations – not SaaS.

The fundamental limitation of conventional integration products (whether hosted on prem or “in the cloud”) is that they are single-tenant. As such, each customer must buy, install and maintain its own copy of the product and must do so at every location where integration is to occur. As a result, using conventional integration products to integrate SaaS greatly increases cost, complexity and time to deploy while also greatly limiting scalability. To date, ISVs have had little choice but to pass that cost and complexity on to the end customer. The underlying premise of this model is that integration is the customer’s problem to figure out.

There is a better way. On-demand integration solutions are themselves built using SaaS technology. Just like SaaS applications, they are a single instance, multi-tenant platform hosted in the cloud. There is only one platform and all customers use that platform for integration. As such, they natively deliver all the great values of SaaS: rapid time to deployment, faster innovation cycles, ultra-low cost and infinite scalability. And they have a dramatically reduced cost structure because everyone leverages one platform.

On-demand integration solutions also eliminate the cost and hassle of maintaining and upgrading each instance of a conventional product. In the on-demand world, upgrades and enhancements to the common platform are continually released to the user community and everyone enjoys the benefits immediately. On-demand integration solutions finally align integration cost with that of on-demand applications – an issue which has been a major barrier to sales for SaaS ISVs.

But most importantly by employing an on-demand integration solution, ISVs can lift the burden of integration off the shoulders of the end customer. On-demand integration solutions allow SaaS ISVs to offer bundled and pre-built integration as a service – solving integration on behalf of the customer. Since the customer is buying the application as a service it only makes sense to buy the integration as a service.

SaaS ISVs have the following options in developing their strategy for integration:

1. Continue to resell conventional integration products and appliances
2. Sell professional services engagements and build custom integrations
3. Build an on-demand integration infrastructure on their own
4. Partner with on-demand integration vendors to package integration for their clients

It’s a classic build/buy/partner decision, but partnering has the advantage of allowing the ISV to stay focused on its core business and to take advantage of an existing on-demand integration solution and get to market quickly with a bundled integration offering.

Forward thinking SaaS ISVs are already moving in this direction. Integration is quickly becoming a strategic imperative and source of competitive advantage. Consumers are flocking to the SaaS paradigm because it saves them time and money, and by providing integration as a bundled part of an application, ISVs can offer a great service to their customers while increasing win rates and accelerating sales.

Integration IS the Application

Integration Is the Application

 

Ok – that statement may be a little over the top but I make it in the same spirit as Sun’s vision for so many years that “the network is the computer.” Using that analogy, I absolutely believe that integration is the application in the SaaS paradigm.

 

I was speaking with the CEO of one of our partner companies last week. He was excited because they had an interesting prospect for their SaaS application and were very close to closing the deal. The prospect liked the functionality of the application and felt it was a good fit for their business. So, the CEO said, all we need to do is integrate them with a few things and we’re all set… 

 

Well the few things turned out to be the prospect’s website, several online trading marketplaces, an on prem point-of-sale system, an on prem financial system, and an online third-party shipping system. Oh and of course the partner’s SaaS application. And please deliver all of that without adding any appreciable cost to my solution such that I get priced out of the deal. Sure – no problem!

 

In the SaaS world of APIs, web services, and interoperability, it is increasingly common to see highly specialized applications focused on solving a specific business requirement but with very deep and rich functionality. That’s the beauty of a service oriented approach. However, this approach relies heavily on the ability to integrate with other highly specialized apps to deliver a complete and holistic solution to the end customer.

 

We get calls almost weekly from SaaS ISVs who have built really ingenious and innovative applications for the Web assuming there would be a way to solve their integration needs. They need to get data from multiple disparate data sources both in the cloud and behind the firewall to make their application “sing” and find that traditional on premise integration software packages or hardware appliances have become a sales and implementation hurdle as well as a long term scaling issue.

 

Integration, and the ability to deliver integration as a service, is quickly becoming a competitive imperative. It’s not enough to just offer APIs to your customers. Increasingly consumers of software as a service also want to buy integration as a service as well. They want a complete and integrated solution that works “out of the box,” can be deployed rapidly, and begins delivering value immediately. When you buy appliances for your house, they come with cords and plugs – you don’t need to be an electrician to use them. Integration is the application…

SaaS Consumers: “Don’t Make Integration My Problem”

The recent acquisition of Cape Clear by Workday signals the beginning of a trend that I think will have profound impact on the SaaS industry.  That trend is consumers of SaaS are beginning to push back on ISVs saying in essence – “don’t make integration my problem.  I am buying a service and I want integration included in that service.”  This has been our view at Boomi since we began building Boomi On Demand two years ago and is why our go-to-market strategy has focused almost exclusively on partnering with SaaS ISVs and technology providers.

The tendency of ISVs has been to sell around the integration challenge in the sales cycle and to base their integration strategy around providing a set of APIs to their application.  Well constructed APIs are essential but not the “Holy Grail” to SaaS integration.  For one thing there are very few SMBs that have the developers to use APIs and companies at the enterprise level would prefer to put their precious R&D resources elsewhere.  Then the tendency has been to sell professional services engagements to build the integrations “one off” which pretty much defeats the benefits of a SaaS offering and becomes an enormous maintenance and scalability issue as the ISV grows.  I first wrote about this issue in an OpSource newsletter last September.

I believe there is growing consensus that integration is the #1 barrier to SaaS adoption and a recent survey by Saugatuck showed that the ability to integrate SaaS with on premise workflows is the #1 criterion of customers choosing a SaaS provider.  And I believe integration is the #1 strategic issue for SaaS ISVs going to market once they have their application built and hosted.

The acquisition of Cape Clear by Workday makes a strong and undeniable statement in the SaaS industry about the strategic importance of integration to SaaS ISVs.  It was important enough to Workday to acquire a company.  I agree wholeheartedly with (former) Cape Clear CEO Annrai O’Toole who said in his blog (http://www.capeclear.com/annrai/?p=29) “…integration is at the heart of hosted applications – and not an on premise, bolt-on like other enterprise vendors believe…”

SaaS ISVs who cannot afford to buy a company have the following options:

1. Continue to put the burden of integration on their clients
2. Sell professional services engagements and build custom integrations
3. Build an integration infrastructure on their own
4. Partner with Platform-as-a-Service or Integration-as-a-Service vendors to package integration for their clients

As I’ve said before, it’s a classic build-buy-partner decision but if integration is not your core competence, I would strongly recommend taking a hard look at the partnering option.

Integration is quickly moving from a non-core offering of SaaS ISVs and technology providers to a strategic imperative and source of competitive advantage.  Forward thinking ISVs will move in short order to plug the integration gap in their offerings.