More businesses are attempting to optimize their operations with the help of analytics, although most of the activity still takes place at the departmental level. Additional value can be gained from cross-functional analytics, but it represents a different kind of challenge because the functional units tend to use different systems and data owners often want to maintain control of their data.

According to recent research by EY and Forbes Insights, 60% to 70% of companies now use analytics at a departmental level, up from 30% to 40% in 2015.

“Companies have had success in one part of the business, they then try to replicate that in other departments,” said Chris Mazzei, global chief analytics officer and emerging technology leader at EY. “The companies that are more mature across a number of different dimensions, those we would put into the “leading” category, are out-performing the others. They’re reporting higher revenue growth, better operating margins and more effective risk management, so there’s at least there’s a correlation between analytics adoption and driving better business outcomes.”

Here are a few things that can hold cross-functional analytics back.

Analytics Isn’t Part of the Business Strategy

Cross-functional analytics is more likely to yield competitive advantages and drive more business value when the analytics are an integral part of the business model and strategy.

“The vast majority of organizations still are not able to say that their business strategy has really reflected the role analytics plays in how they’re trying to compete,” said Mazzei. “There’s this fundamental misalignment that can occur when across the leadership team is not able to have a consistent view of where and how analytics is making the biggest impact on the business strategy.”

Operating Models Don’t Facilitate Cross-Functional Analytics

Executing an analytics strategy at a departmental level such as finance or marketing is relatively easy because it’s clear that resources need to be dedicated to the effort. When it’s a cross-functional endeavor, who’s responsible for providing, funding and managing those resources? What should the data flow look like and how can that be facilitated?

“If you’re trying to deploy analytics across the organization, the operating model becomes much more important,” said Mazzei. “Do we have a centralized team? Do we distribute analytics resources in the individual business units or functions? What’s the relationship between those teams?”

Like bimodal IT, bimodal analytics services benefit the enterprise and the departments simultaneously. The centralized group helps facilitate best practices and ensures appropriate governance while dedicated resources tend to have specialized knowledge of that particular function and its analytics requirements.

The Initiatives Aren’t Designed Well

Analytics efforts should drive business value. There’s a lot to do, but not everything will have the same level of impact or necessarily achieve the desired results, so the desired business outcomes should drive the prioritization of analytics efforts.

“Initiative design is really important and are there competent frameworks/processes you use for that,” said Mazzei.

Not surprisingly, companies are still at very different stages of maturity in terms of having any kind of consistent process for designing an analytics initiative. The more analytically mature a company is, the greater the likelihood is that they have common frameworks. There is also a common understanding of what the term, “analytics initiative” means and common tools for executing that, Mazzei said.Analytics Isn’t Part of Business Operations

As companies embrace analytics and mature in their use of analytics, business processes tend to change. It’s wise to think about that and other impacts early on.

“The more mature companies are thinking about that earlier in the process and using an initial point of view about what that intervention needs to be to inform how you design the analytics themselves,” said Mazzei. “A lot of companies don’t think about that early enough.”

According to the report, design intervention is “Translating all the upfront goal-setting, modeling, and methodology into action— making analytics insights an integral part of business operations.”

The True Value of Analytics Isn’t Understood

Interestingly, analytics enables organizations to measure all kinds of things and yet success metrics may not have been defined for the analytics initiatives themselves.

“That really matters because [if] you can learn, what’s working and what’s not earlier on, you can change the nature of the intervention or the analytic you’re building,” said Mazzei. “It’s that feedback loop you have in place.”