Modernizing IT is meant to rationalize processes, simplify operations and reduce friction. When systems resist integration, it can also lead to frustration and cost overruns.
Organizations cautiously loosened the purse strings this year, as the macroeconomic environment improved. But investing in digital transformation remains an enterprise priority, according to two reports from analyst firm Forrester.
More than 4 in 5 IT decision-makers expect their organization to increase tech spend heading into 2024, according to the Forrester studies, published Aug. 1. Nearly the same proportion anticipate increased investment in cloud infrastructure and development services.
“Cloud is the essential element of the IT modernization story,” Forrester Principal Analyst Lee Sustar told CIO Dive. “But it's not the only element. People are trying to modernize around [their] legacy applications because they can't replace them.”
Cloud, now classified alongside the on-prem data center as a “traditional” technology by Forrester, has different meanings based on enterprise priorities and the state of the existing tech stack. Efforts to avoid vendor lock-in, for example, or the desire to find the best cloud for a particular job often add unwanted complexity to tangled IT ecosystems.
“People who want multicloud in order to mitigate vendor concentration risk may be willing to accept a degree of complexity,” Sustar said. “Having said that, they want to make sure that the workloads they're putting in cloud are in the right cloud for the right price. Or, maybe it's good enough to run on the data center as it has been for years.”
Rather than adding more cloud, Forrester recommends focusing on interoperability, deploying capabilities to connect existing environments more seamlessly.
Moving data, a cornerstone of cloud interoperability, can be costly.
The cost of shifting workloads between two cloud providers is an issue that has reached beyond the C-suite to the halls of corporate justice. The Federal Trade Commission and British regulators have opened separate investigations into anticompetitive practices in the cloud market, including hefty fees.
Some companies are willing to pay the price, Sustar said.
“Interoperability can be fairly limited, given egress charges,” said Sustar. “But there’s a well-known pattern with AI of training a model on [Amazon] SageMaker and then shipping it over to Azure, where Microsoft Power BI can be used for insights.”
While there are a growing number of managed service provider solutions aimed at simplifying hybrid ecosystems, frictionless interoperability between different cloud providers and with on-prem systems remains elusive.
Partnerships that connect data storage and hyperscaler platforms, like deals Snowflake has with Microsoft and Google Cloud, offer a partial solution, as does the data integration agreement announced by SAP and Google Cloud earlier this week.
Sustar points to the BigQuery connector linking Google Cloud and Microsoft Azure as an example of a promising hyperscaler integration and a potential sign of future collaborations.
“The cloud providers are looking at other clouds that customers might want to use as something to be partially integrated into some of their core management plans, but not extensively,” Sustar said. “Users are pushing for that, but it's not there yet.”