As organizations become increasingly digital, there has been greater collaboration between IT and the business -- a necessary partnership when it comes to enabling successful data analytics.
Modern IT departments need to be data-savvy and have strong collaboration skills. They also need to ensure the IT ecosystem can support real-time or near-real-time insights, which includes ensuring the timely flow of reliable data.
IT's role is increasingly become the enabler. IT departments enable analytics and data use in organizations by putting processes in place, as well as installing the data platform and analytics tools that best fit the organization's needs.
"It is necessary to strike a balance between self-service and governance … by putting in place data architectures that support both," said Myles Gilsenan, vice president of data, analytics and AI at enterprise application services company Apps Associates.
The more sophisticated IT shops are oriented toward data, as opposed to oriented just toward infrastructure, Gilsenan said. Such organizations may advocate for things like data catalogs to help the business understand and properly use data assets, including data in raw form.
Necessary skills to enable data analytics
Modern IT organizations need a mix of infrastructure and data skills because both enable an insight-driven enterprise. Infrastructure skills are necessary to achieve an architecture that can support data analytics requirements, including scalability, data governance and data security.
Today, the building, operating and innovating of the value proposition using AI is getting simpler with the aid of advanced AI cloud platforms.
"This means we will need business, product and technology teams who bring skills with deep experience in leveraging data and to offer comprehensive products and capabilities that are aligned to the business context," said Rizwan Akhtar, executive vice president and CTO of business technology at real estate services company Realogy Holdings.
Fundamentally, IT needs to have stronger math skills, including linear algebra, statistics, calculus and maybe inferential geometry -- skills which data scientists tend to have. However, given the mainstream adoption of AI and machine learning (ML), there are now tools that make it easier for non-data scientists to do more.
Augmented analytics is a great example of that. However, someone needs to understand how AI and ML work. Similarly, there needs to be someone with infrastructure expertise that understands the nature of workloads.
IT has multiple methods to enable scale-out. One such method is a mixed infrastructure consumption model. Key to these methods is inclusive of capacity in both on-premises and public cloud environments and the use of scale-out technologies, such as hyper-converged infrastructure, which enable elastic use of storage, compute and network services.
"Resiliency -- the ability to withstand issues by making use of optional routing, standby capacity and fault-tolerant services -- is core to the IT operating model," said Wendy Pfeiffer, CIO of IT cloud software provider Nutanix.
There is also a need for business domain expertise, which tends to involve collaboration with the business to understand how to use AI and ML to solve business problems, such as understanding customer journeys.
Some organizations have centers of excellence, which aggregate best practices throughout the organization, complemented by remote experts that work in a department or line of business as specialists.
Meanwhile, shadow IT is still alive and well, fueled by analytics vendors cashing in on departmental purchases. Over time, IT departments tend to realize there is more than one analytics platform in the organization, and many instances of the same platform exist without the benefit of volume discounts.
An emerging trend is for IT and data scientists to enable citizen data scientists. This makes it easier to set corporate standards for analytics in terms of capabilities, data access, data hygiene, data integration and data governance.
In addition, the IT department has visibility into the platform's use in the organization. Without IT's involvement, it may not be clear whether a platform is the right platform or whether the data is reliable enough to produce reliable insights.
Data analytics necessitates new roles
Many companies want to employ data scientists, though that role is not entirely new. In the past, the role may have been statistician or actuary -- one who deals in probabilities.
Not all statisticians have programming skills, like the typical data scientist. In fact, a sign of market immaturity has been the race to hire data scientists before understanding what they need to be successful, such as collaborating with the business instead of working in a corner of the IT department.
A new role is the data engineer, who helps ensure data quality and the data pipelines that are necessary to enable analytics. Quite often, these professionals have held other IT positions, such as systems administrator or database administrator.
"Data engineers are folks with strong knowledge around extract, transform and load tools, varied programming languages and database systems. Data scientists seem more centered around multiple programming languages, specifically targeted toward machine learning," said Jennifer Fahey, principal consultant at enterprise cloud solution provider Syntax.
Data analysts are core to the process because they have more of the business knowledge but may or may not have basic programming skills. They may reside within a business unit, work on a data team or both.
Business analysts are still in the picture. In fact, they're the most likely people in the organization to become citizen data scientists because they're more technically oriented than other business professionals.
The C-level roles -- chief data officer (CDO) and chief analytics officer (CAO) -- may report to the CIO or not. In some cases, the CIO may become a CDO or CAO in a digital business that prioritizes analytics.
"Data teams embedded in business units is the only practical way to keep up with the hypergrowth of data sources and the data demands of the business," said Satish Jayanthi, CTO and co-founder of data transfer tool provider Coalesce. "While IT enables and plays a vital role in creating governance policies, [the] business will embed data teams who will be responsible for owning and sharing high-quality data."
Data teams and IT maximize data analytics efforts. IT teams continue to play a pivotal role in enabling organizations' analytics operations, verifying the right platform and checking the reliability of data to produce insights as organizations ramp up data use.