Integration continues to frustrate organizations trying to use SaaS applications. TechTarget reports that application integration showed up as a top cloud problem in two surveys it recently conducted.
You may recall earlier this year a KPGM survey that found 31 percent said integrating cloud services with their on-premise applications and systems was more complex than expected.
TechTarget’s Cloud Pulse survey found similar results: 34 percent said SaaS programs can’t interoperate with other programs from the cloud or in house.
“And once again, customization was tied with integration: Even with customization, 34 percent of Cloud Pulse respondents said SaaS apps can be inadequate for the client's business needs,” reports SearchEnterpriseLinux.com Site Editor Jan Stafford.
Since we have iPaaS and other options for integrating the cloud, you have to wonder what’s going on. Cloud computing and integration expert David Linthicum recently offered a reason for some of that disconnect (ahem).
“Data integration, no matter if you’re integrating internal systems or those in the cloud, is typically an afterthought,” he writes. “Companies now spend millions of dollars to migrate systems to the cloud without a clearly defined plan as to how that data will be integrated with existing enterprise systems. This makes the cloud project much more risky and costly than it has to be.”
It’s not the cloud that’s the problem, he adds, but it does highlight a persistent issue in enterprise IT: a lack of integration skills, including the ability to plan an integration strategy.
And then there’s this: The dynamic resource assignment environment of cloud and SaaS makes integration more difficult, she says, citing Cloud Pulse Survey respondent Saurabh Sharma and Tom Nolle, an Ovum senior analyst and CIMI Corp. cloud consultant.
Nolle told her that it’s harder to achieve application and data integration when things move — and they do move in the cloud.
“Also, Web services application program interfaces, or APIs, aren't the promised silver bullet for clear interaction between SaaS and on-premises applications,” Stafford writes. “That's because on-premises apps have been developed under different standards and often need much custom-code development to interact with SaaS environments.”
So, what’s the solution? Stafford lists several approaches that offer hope for resolving this very stubborn problem of integration, including:
Linthicum warns, though, that a long-term solution to integration’s challenges — and costs — requires rethinking how you handle integration.
“The solution to all this is to go back to the basics. Those who are moving to cloud-based systems need to review their existing data integration strategy and determine the best way to integrate existing systems with cloud-based data,” Linthicum states.
“This also means selecting and deploying the right technology for the job.”
Talk about having your head in the cloud! You would think companies would’ve learned by now, but it seems many organizations thought cloud was their ticket out of hassles like forethought and planning. That means they’re paying a premium for hand-coded, custom solutions — a lesson they should have learned with on-premise integration. The difference, as both Linthicum and Stafford notes, is that integration only becomes harder with the cloud.
Organizations are in denial when it comes to integration. But that’s just poor budgeting. It’s time to wake up and tackle integration as a primary issue, rather than an afterthought.
If you’re interested in funding a more comprehensive approach to integration, you might find this checklist of six ways to calculate cloud integration ROI helpful.