There’s a cloud computing question among IT and business leaders that has been around for a while: How much cloud flexibility and choice is enough, and how much is too much? This year, not surprisingly, it has taken on heightened importance with the acceleration of workload migrations to cloud and application modernizations.

Answering this critical business and IT operations question is a bit complicated. But, by taking an informed, strategic and balanced approach, businesses can successfully answer it.

In theory, having an abundance of choice is a great thing. The reality with cloud computing, however, is that teams must trade off portability with tight integration, and focusing too much on portability can lead to wasting precious personnel resources, time and money with unnecessary abstraction layers and inefficient use of cloud resources. As the chief cloud architect for a technology vendor, I spend a lot of time with customers, partners and prospects delving into the details to find the proper balance, and I’ve learned a thing or two. Here’s how I would go about answering this question.

Multicloud is part of the answer, but it’s not the answer.

Public cloud vendors love to tell customers who don’t want to be locked into their ecosystem that the vendor’s hybrid or multicloud management solution can be a unifying layer, giving freedom and choice back to the customer. And to a certain extent, it can. But these solutions are typically targeted at infrastructure management, not application design and operation. In my experience, the complexity involved in running part of an application in one cloud and part in another rarely overcomes the benefits, and trying to build an application that can be deployed in more than one cloud simultaneously usually involves a lowest-common-denominator approach to platform use that leaves many of the benefits of cloud still on the table. Multicloud applications are rare and difficult to justify, no matter what the vendors tell you.