You are here

You are here

Cloud app integration: Take the road less traveled

Peter Merkulov Chief Technology Officer, Globalscape

The race to digital transformation is like a marathon without a finish line. The ongoing journey is a course that frequently changes. As new technologies and trends emerge, the digital transformation journey will challenge an organization's strengths, expose its weaknesses, and test its endurance.

While these new applications and solutions promise more innovation and efficiency, the sheer number that's available creates unavoidable complexity, especially when it comes to managing the data and various systems within a modern IT infrastructure.  

As with any innovation, success with iPaaS isn't instantaneous. While iPaaS is still in its infancy, application integration and data integration are not. Most organizations haven't developed a mature integration strategy that incorporates iPaaS because cloud integration is only now becoming a pressing issue.

With new stakeholders involved, strategy becomes more complex. Without a careful strategic approach, businesses can experience process inefficiencies, piecemeal sourcing of applications, and/or duplicated efforts within the organization. It's quite easy to overlook important steps within the process or unknowingly overlook vital considerations.

To avoid getting lost in the clouds, consider these six points.

1. Evaluate building versus buying

This is an important first step that should require long-term thinking versus impulsive decision making. The integration, whether built in-house or bought from a third party, should fulfill all of the business requirements and serve as an efficient use of time, energy, and money. Decision makers must consider the route that will protect both line-of-business objectives and the long-term sustainability of the business.

Some central questions that should be considered are:

  • What will it really cost to build or buy?
  • Do you have the in-house expertise to anticipate or forecast possible problems in building integrations?
  • Do you have someone who can design integrations that will scale up (the number of users) and scale out (the number of programs)?
  • Would IT have to give up other priority projects to build the integration?
  • Would IT have the resources and skillset to maintain and continue to improve the custom code once the integrations are built?

2. Get on the same page

Problems that arise don't usually come from the platform itself but from misunderstood stakeholder requirements. Be sure that you've spoken with all involved parties and that everyone understands the expected output of the initiative. Getting on the same page can be one of the most challenging aspects, which is why it should be addressed head on and early within the process.

3. Assess integration patterns

There are a number of integration patterns to consider in an organization—application integration, data integration, B2B integration, reinforcement of business workflows—and each requires its own set of best practices. Don't overlook where these applications are located—on-premises or in the cloud—or whether you might need support for hybrid or Internet of Things (IoT) integrations.

4. Don't forget legacy tech

The trickiness with legacy technology is that it was adopted for one objective but is now being modified for another. Without a sound understanding of the system and the new requirements, the transformation can be a painful process, fraught with unexpected issues. A careful audit of these systems ahead of time and proper project scoping can mitigate some of these risks.

5. Accommodate existing integrations

While it's natural to think that the rip-and-replace method is often the best, in the case of existing integrations, it might not be. It's likely that your organization has spent a lot of time and money making existing integrations work together.

As the old saying goes, "If it ain't broke, don't fix it." You should consider ripping out or modifying current integrations only if they aren't serving your current goals—and if they aren't, be sure to understand why.

6. Future-proofing

You've thought outside the box, followed best practices, and your integrations are now in place—which means you're done, right? Well, maybe not.

For one, integration efforts are only as stable as the systems they connect. When a cloud app updates its APIs, there are ramifications for your integrations. Avoid getting stuck on an "API treadmill" that requires you to test integrations with each new update by mapping your integration tools to these specific apps/integrations. 

If you are doing ad hoc integrations, it's best to make sure that the tools being used are ones that make it easy to spot and remedy issues.

You also shouldn't feel comfortable using a new tool without properly addressing security. Since iPaaS connects to a large number of systems, security should be a primary concern. The ability to pass security audits and validations is the minimum security requirement.

Permissions and access controls powered by IT are also important, though often de-emphasized. For a trusted platform that services multiple cloud applications and lines of business, standard controls should include full security at the credential (login) and workflow levels, secure groups of workflows, role-based authorization with multiple user types (admin, read-write, read-only, etc.), and encryption at rest.

Your business needs, in addition to the technology you're using, will constantly evolve. Don't be a victim to the set-and-forget mindset. It takes a lot of work to reconfigure a homegrown solution. So instead of being caught in that dilemma and eventually hitting a wall, make sure that your connections and tools have the potential for scalability and fine-tuning.

Why now? Lots and lots of cloud

The average enterprise today uses more than 1,400 different cloud applications, according to Skyhigh Networks. While such arrangements certainly offer cost and productivity advantages, having multiple places where data and applications live can exacerbate complexity.

Making matters worse, the pace of innovation isn't slowing. This can create some anxiety for those responsible for managing the onslaught of new tools, as advancements begin outpacing the necessary points of integration.

For decades, businesses have had to break down legacy data silos and increase integration. Building data connectivity between various mission-critical cloud applications is where progress still needs to be made. Traditional silos of on-premises networks have just moved to the cloud, rather than being broken down with new processes and technology.

Fortunately, with the advent of integration-platform-as-a-service (iPaaS) technology, enterprises are one step closer to solving modern connectivity and efficiency problems.

Industry analysts say the iPaaS market is projected to hit $2.7 billion by 2025. While the business benefits are undeniable—streamlined workflows, increased business agility, centralized control, self-service capabilities, and more—there’s no one-size-fits-all solution, and the devil is in the details.

Stay on top of it

As organizations work to define what digital transformation means to them, integration strategies and tools will continue to evolve. Speed bumps and extra turns might slow your process, but ongoing efforts to educate yourself and your teams about common missteps will get you miles closer than you were before.

Keep learning

Read more articles about: Enterprise ITCloud