Cloud computing has proven its value to enterprises and service providers alike. On the business side, the cloud represents potentially lower costs thanks to pay-as-you-go pricing and the option to scale easily as the user base grows. For service providers, cloud computing keeps customer acquisition costs low, translating into higher operating margins. Based on these realities, it would seem that cloud computing is a pure win/win for all involved.
However, the actual process of migrating enterprise data to the cloud exacerbates a problem that IT has long faced: Before the team can determine which applications to move to the cloud, it needs a complete understanding of its assets, as well as the ability to connect that knowledge to related information about contract terms, licensing and usage patterns. Gathering that information is more difficult in today’s enterprise than one might think. There are three reasons for this:
1. The naming patterns don’t align across systems. Perhaps your IT asset management system notes the presence of Dreamweaver, which was made by Macromedia, which was acquired by Adobe. Your operational system might note the same product as simply “Adobe.” Multiply this nomenclature issue across multiple systems, vendors and products, and IT ends up with a difficult, manual reconciliation task that can significantly delay cloud computing initiatives.
2. IT can’t get an accurate accounting of usage patterns or cost. Enterprises make cloud decisions based on cost and usage data. The naming pattern challenge noted above makes it difficult to get that data. There are additional challenges here; enterprises need to understand the business lineage of software and bundling options in order to aggregate and analyze data across systems, as well.
3. Good business decisions require accurate market data. In addition to assessing internal information, IT needs to know whether the enterprise’s existing software is available as compatible cloud offerings, which licensing alternatives will deliver the most cost-effective migration, and where the enterprise is in the support lifecycle.
The above challenges require enterprises to adopt a common language of IT a taxonomy that categorizes information so it can be used to normalize data and support informed decisions about cloud migration. Once that is done, enterprises can evaluate the two kinds of cloud applications: those that are loosely coupled to other data within the enterprise and those that are tightly tied to enterprise data.
Salesforce.com’s sales force automation application is an example of a data set that is loosely coupled to other assets in the enterprise. Most of the data on which salesforce.com relies is in the cloud, with only a small amount elsewhere in the enterprise. Therefore, enterprises can make the connections between relevant data with relative ease.
By contrast, an application like ServiceNow hosts large amounts of data in the cloud but also relies on multiple enterprise systems. Such applications need to be able to assimilate data from any type of system, filter it to the most relevant set, and make it fit for consumption by the cloud application. The application must be capable of performing all of these actions on an ongoing basis. That requires extreme processing of data as well as a common language of IT.
To simplify migration and rapidly scale cloud application adoption, enterprises should start by addressing these common hurdles. Before they migrate, they must solve the data disparity issues in their application landscapes. When they do so, IT teams will make better business decisions about what to keep on premise and what to move to the cloud, and they’ll lower the cost and risk of migration.