The Inovis Blog

Image Source: http://beckerinfo.com

Erik Huddleston, the Chief Technology Officer at Inovis, was recently asked an interesting question that he enjoyed pondering. The question was:

In ten years, will *every* Cloud provider, SaaS vendor, ERP vendor, et al, include enough integration capabilities such that having independent integration as a services is simply unnecessary?

Erik responded:

Let’s break the discussion down into two different questions:1) In house vs. Cloud
2) In app vs. External

For the first question, I believe you will continue to see both. For the second question, I think you will continue to see external services be the dominate mode instead of “baked into the app integration” (excluding OEM or private labeled/integrated IaaS to Cloud).

Why do I think that? History.

The complexity of the business processes currently being integrated the diversity of integration patterns emerging continues to accelerate. Until that stagnates, it will continue to be a specialist market.

Likewise, if you limit it to just 10 years, I think it is a virtual certainty that integration will remain independent and will continue to increase in prominence. SaaS is driving a second wave of “best of breed” in the enterprise space. You can get very specific use cases satisfied in a point business process cheap (opex with low implementation costs), fast (no hardware to procure and set up, no IT lines to wait in to get it installed), and easy (bypass IT entirely and solve your business problem). Just like the last best of breed wave in the 90s, we will create tons of islands of information…an environment ripe for integration. Just like the last wave, those platforms will be used as an extensibility point to add visibility, reporting, workflow, and lightweight business process functionality that the SaaS app doesn’t support “out of the box”. This takes on greater prominence given that the SaaS model removes the traditional customization and “dirty” integration models of enterprise software such as direct DB access and custom code modification.

The other interesting SaaS trend that will continue to drive cloud integration is the proliferation of cloud based custom application platforms. If you look at the innovation in Google Apps and the Amazon/IBM partnership, the cost, time, and skill level required to get a completely unique enterprise app launched is rapidly being driven down. So cloud platforms should significantly increase the number of custom apps in the enterprise, not decrease them. It will be tough to leverage a “baked in” integration technology from say, Salesforce or NetSuite to orchestrate a complex integration between three of four SaaS apps.

There is and will continue to be significant enterprise architecture and executive pressure to centralize integration in the cloud. The CIO and CFO will use cloud integration platforms as a single point of governance, risk management, compliance, enterprise visibility, and exception management. Those are tough functions to string together between each system individually, regardless of the robustness of integrated capabilities.

As to in house vs cloud, there will be a market for in house integration software for a LONG time (well beyond 10 years). Much like the application space, the unique advantages of the SaaS model will mean that cloud integration platforms will significantly outstrip “in house only” vendors in terms of features and unique capabilities. That said, cloud services will never have the one feature that will dominate many buyers decision process: “does it run behind my firewall?” That gives a significant advantage to hybrid platforms.

I’m curious to hear your thoughts. What do you think of Erik’s response? How would you answer this question?