2022 Analytics Trends

2022 is the year in which you must get your reporting in order. Many companies now find themselves in a mess of reports and the forces that make this worse are often stronger than those pushing for consolidation.

Reports and dashboards have often been proclaimed dead. But we see that they are very much alive. Sure, they are by no means the silver bullet for all information needs in a company, but they are the undisputed backbone of corporate information distribution. Reports and dashboards done right help quickly make sense of data and drive determined action to create value.

According to our worldwide BI & Analytics Survey, 84% of companies use standard reports, and 78% use dashboards or interactive analytical apps. These figures depict the use of reports and dashboards in a distinct tool while companies on average have more than three analytics and BI tools in use.

Information presented in a standardized form offers a low entry barrier for all users. It helps to satisfy repeated information needs. There is a substantial need for serving information to business users in a standardized way.

Now, while we clearly rely on reports and dashboards to direct our actions, corporate reporting is often in a terrible state. Many companies suffer from having too many reports in and outside their analytics and BI systems. Your company probably suffers from that overabundance too. Having too many reports seriously undermines trust and diminishes analytics and BI efficiency. Furthermore, it undermines your efforts to get value out of your data.

“Having too many reports seriously undermines trust and diminishes analytics and BI efficiency.”

In practice, it is always easier to request or build new reports than to have them deleted or decommissioned. This force is a major cause of the reporting mess in companies.

Changes in business, the competitive landscape or customer behavior as well as changes in management or corporate structure all magnify requirements for new information and reports. Further pressure is exerted by more data sources delivering more data to more users in more departments through more analytics and BI tools to support a growing variety of use cases.

While these forces are manifold and powerful, there is little evidence to suggest their pressure will diminish. They clearly show that analytics and BI must deliver information quickly to support these new requirements. The required time-to-insight can only be achieved when allowing and supporting self-service, which poses further challenges to the reporting environment.

The bottom line is: Whatever happens, more reports get created. And the more reports a company has, the lower the share of what is used. It is not too uncommon to find that 80% of the reports available have not been run in the last twelve months, especially in self-service environments lacking the required conditions. This is a huge waste of resources and a huge cause of distraction for decision-makers and employees at all levels.

The reporting mess is a significant cause of dissatisfaction with corporate reporting. Discontent arises from the high effort in creating and maintaining these reports, especially if a large proportion are only ever used once or twice. Too many reports result in inconsistencies, and this and they are the gas that gets lit by sparks of distrust in the information present. An overabundance of reports also draws the attention of decision-makers away from what matters most and blocks the resources of front-line workers from serving their customers. While the effort to create and maintain a massive number of reports is huge, it often does not pay off. So, you must clean up the reporting mess or even prevent it in the first place.

Constant change in a covid economy has companies thinking differently about their data. New analytics use cases continue to emerge. Digitalization has accelerated. And data exchange continues to expand—among customers, suppliers, and partners.

Driven by the need for flexibility and agility in delivering new projects along with the growing number of cloud-native data sources in the enterprises’ ecosystem, data and analytics have continued to move to the cloud. As a result, we have seen the emergence of new architectural designs such as data mesh, data fabric, and data lakehouses. These concepts all attempt to reconcile the growing complexity of cloud data collection. Where raw data is collected without any consideration for how it relates to other data—if it can be used or trusted.

"Maintain flexibility and agility in an increasingly complex data landscape with a ‘trust but verify approach.”
 

All these architectural designs have their merits, as they highlight the growing complexity of the data landscape. Nonetheless, this complexity is challenging business users and analysts who were hoping that self-service alone could address their daily analytical needs. Even for savvy data users, making meaningful use of the data has become more difficult. As a result, having trusted shareable and reusable data in support of analytics has become even more important.

The data management practice alone can’t address the full spectrum of analytical use cases with the flexibility and agility needed by organizations even with increased investment in activities such as data catalogs or augmented data integration capabilities, and in dedicated resources like data engineers.

As a result, organizations need to adopt an adaptive analytical governance approach built on a ‘trust but verify’ methodology to overcome data complexity. To do so organizations need to articulate various rings of analytics governance that fit the business requirements, skills, and needs of the various users across the business:

• Inner ring for the centrally governed analytics representation of data.
• Middle ring for department or BU-specific analytical governance.
• Local ring for users who will need to augment analytics from central or departmental governance.

Analytical governance differs from data governance as the former is based not only on data, but it also includes the definition of calculations, metrics, AI/ML models. Basically, it’s all of the analytical content and all the metadata relating to the content itself like the freshness of the analytical content, author, etc. Analytical governance is powered by a rich underlying metadata layer organized in the form of a semantic graph.

The trend toward cloud analytics is clear. One-half of the participants in our research currently use cloud computing products for their analytics and data processes, and three-quarters expect to use the cloud for these processes eventually.

Our research has also shown that 86% of organizations expect the majority of their data to be in the cloud eventually. Part of what is driving the adoption of cloud analytics is that many organizations are already operating in the cloud. More than one-half (59%) of the participants are analyzing data from cloud applications. They have experienced some of the benefits of cloud applications and it’s natural to extend those benefits to their analytics.

However, it’s not all smooth sailing. Organizations still report a slight preference for on-premises analytics at a rate of 39% compared with 34% that prefer cloud-based analytics. The research shows that more than one-half (57%) of those who are not using the cloud are concerned about security issues. Clearly, analytics software vendors and hyperscalers need to continue to address these issues by providing secure and trusted platforms, ideally with certifications to meet different organizations’ requirements.

“Containerization combined with a microservices architecture creates the scalability needed to support large numbers of users and large volumes of data.”

Interestingly, those organizations that prefer on-premises deployments report higher levels of satisfaction than those preferring cloud deployments. Those preferring on-premises deployments also report that their analytics technology has had a greater impact on improving their activities and processes. Perhaps these differences result from the fact that some on-premises offerings have not been fully migrated to the cloud yet.

Therefore, these particular cloud offerings are lacking some features compared to their on-premises predecessors. The research also shows two-thirds of organizations (65%) that prefer on-premises deployments have a hybrid of on-premises and cloud infrastructure. Perhaps some of the reduced satisfaction results from challenges interoperating between on-premises and cloud deployments. Further, the “cloud” is not one thing. Almost one-half (42%) of the organizations in our research are using at least two different hyperscalers. Consequently, organizations should strive for hybrid analytic architectures that can span these on-premises and multi-cloud configurations.

Containerization provides the means to achieve this deployment flexibility. Containerized applications enable the same code to be deployed on-premises and on the various hyperscalers’ platforms. Containerization combined with a microservices architecture creates the scalability needed to support large numbers of users and large volumes of data.

The other reality of making cloud analytics work is avoiding cloud cost surprises. Organizations prefer predictability in the costs associated with analytics software. Subscription-based pricing models vary from vendor to vendor; some are priced on simple metrics such as numbers of users, while others use more complicated metrics based on different usage parameters such as the amount of data processed, or the number of reports processed. If the pricing metric is not easily understood and predictable from one billing period to the next it can lead to some very unpleasant surprises when the bill arrives. Vendors can also make cloud costs more predictable by offering tools to monitor and manage the usage throughout the organization.

Notwithstanding the potential issues associated with cloud analytics, we expect to see their popularity continue to grow toward eventually eclipsing on-premises deployments. But it’s wise for organizations to have their eyes wide open so they can understand and manage their cloud deployments. The cloud offers many benefits and organizations should seek to maximize their impact with a careful and thoughtful selection of their cloud analytics vendors.