Artificial Intelligence and DaaS

The Six Biggest Challenges Plaguing Your Enterprise Intelligence Efforts

From data silos to organizational bias, the future of intelligence has several challenges to overcome. Explore the largest with IDC's Dan Vesset.
Pinterest LinkedIn Tumblr

We’ve discussed the need to redefine enterprise intelligence and what the future of intelligence will look like for technology and business leaders, and we know that organizations are paying attention. In fact, enterprises spent $200 billion on data, analytics, and AI software, hardware, and services last year. That doesn’t even include the investment in external data and internal labor costs to further fuel intelligence initiatives.

But despite all this awareness and investment, the question remains: why are enterprises struggling to develop the capabilities that drive the future of intelligence?

Silos: The Obvious Barrier to Intelligence

Perhaps unsurprisingly, organizational silos are the largest challenge to enterprises trying to develop their intelligence capabilities. Often, they have access to data and the analysis of that data in a silo or through a very narrow point of view and are incapable or unwilling to apply the resulting knowledge gained outside of the silo or project.

Silos don’t just exist in organizational structures. Silos can be mindsets, project layouts, and strategic elements as well. Here are just a few ways that silos impede intelligence efforts and prevent learning at scale:

  • Outlook Silos: Organizations stumble when they focus most of their efforts on information capture and report delivery that highlight past performance, rather than supporting all the steps in a decision-making process that involves a range of descriptive, diagnostic, predictive, and prescriptive analysis methods.

    Enterprises should instead strive to provide pervasive decisioning across projects and departments to evaluate scenarios, understand key drivers and recommend next best actions in a collaborative environment among internal and external subject matter experts, both humans and machines. Data and knowledge gained from analyzing data often has multiple uses across the organization, apart from the original team or project that generated them. Foster an environment that allows the entire organization to learn and benefit.
  • Scope Silos: Organizations today can’t see the intelligence forest for the trees; they approach projects as functional business process silos or even subprocess silos. Data generated for or by that individual business process remains there; for example, customer experience management data is not connected with service fulfillment data. These are discrete business processes, but both could benefit from the other process’s data or knowledge. This silo issue has existed for years but is becoming worse in today’s data-rich world.

    Scope silos extend to more than data. They also affect organizational structure, staff, and processes. They result in lost productivity, exposure to unnecessary risk, opportunity cost, and in substandard customer, employee, and external stakeholder experiences – all challenges that intelligence projects are meant to overcome.
  • Technology Silos: Technology silos, while issues in and of themselves, also prevent organizations from addressing outlook and scope silos. Without a cohesive and comprehensive data, analytics, and AI technology architecture, organizations can’t see recurring patterns or correctly synthesize internal and external data sources into information and then into knowledge. They are too slow in their decision making, vulnerable to substandard strategic and risk management practices, and unable to identify and address cross-functional opportunities.

Since silos are so pervasive in enterprises, they seem impossible to overcome. However, organizations are already starting to address these issues, especially those who are seeking an advantage in the digital economy. IDC predicts that by 2022, 75% of enterprises will embed intelligent automation into technology and process development, using AI-based software to discover operational and experiential insights to guide innovation – and begin to knock down some of those silos.

The Hidden Challenges to Enterprise Intelligence

While silos are certainly the largest hurdles to overcome in the path to greater enterprise intelligence, there are more subtle, insidious barriers to success. These are cultural challenges that are built into the very fabric of how organizations operate. These three challenges all relate to how enterprises approach and address data in their organization:

  • Data Aversion: also known as the ostrich effect, data aversion occurs when the organization’s culture rejects valuable information in favor of “gut feel”. Ignoring the knowledge gleaned from data or refusing to even analyze data in a meaningful way opens the organization up to all sorts of biases in decision making that could be avoided. A great example is confirmation bias; this tendency to interpret new evidence through the lens of validating one’s existing beliefs or ideas can be dangerous in organizations. Fostering a culture that values data and data literacy can help combat people-created biases, but a strong code of ethics is also needed for data protection and intelligence technology usage as well.
  • Data Illiteracy: When enterprises don’t prioritize building competencies around understanding and communicating about data, data illiteracy thrives. Data illiteracy prevents the entire community from using data, and it also leaves employees and leadership vulnerable to being unduly influenced by misinformation or information pollution. Organizations need to create a common language around data and should employ the help of both internal and external subject matter experts to foster this common language and to build data literacy throughout the organization.
  • Lack of Data Intelligence: Understanding how data is collected and works is one thing; understanding how to correctly analyze that data and apply that knowledge to solve problems is another. Data intelligence is required to successfully turn data into intelligence within an organization. Lack of data intelligence further fuels distrust in data, information, and even insights derived from that data. A lack of data intelligence extends that distrust to AI models and projects that depend upon the data and can derail an organization’s digital transformation.

Tackling Organizational Challenges to Enterprise Intelligence

They may sound insurmountable, but organizations can defeat these challenges to greater enterprise intelligence. Executive leadership needs to foster an organizational evidence-based culture, encourage high data literacy among its employees, and support practices where information shapes decision making by both people and machines to combat data aversion, illiteracy, and lack of data intelligence.

Additionally, organizations need to source and promote closer collaboration between internal data scientists, data engineers, data architects, business subject matter experts, and external networks and experts to move data and information out of silos and encourage delivery of insights at scale.

The first step on the path to enterprise intelligence for many organizations is to assess their data and their processes to see where silos and cultural challenges are blocking intelligence capabilities from flourishing. Learn how to map your organization’s data and processes with IDC’s new eBook, “The Future of Intelligence: The Business Leader’s Guide to Mapping Your Organization’s Enterprise Intelligence”.

Dan Vesset

Group Vice President, Analytics and Information Management