IT Decision-Making Through the Lens of Gartner’s Analytics Maturity Model

An IT business analytics solution’s capabilities should be evaluated for its ability to aid in proactive decision-making. The Analytics Maturity Model (AMM) has its roots in the software capability maturity model (CMM). Both models describe the different stages companies travel through in order to reach process maturity. Whether an IT team is analyzing their current solution or evaluating a new one, utilizing Gartner’s four stages of analytics maturity model provides a valuable perspective.

Gartner ranks data analytics maturity based on a system’s ability to not just provide information, but to directly aid in decision-making. More mature analytics systems can allow IT teams to predict the impact of future decisions and arrive at a conclusion for the optimal choice.

Gartner’s four stages model of data analytics maturity can help with both assessing the current state of IT business analytics systems and revealing an optimal path forward. The four stages, in order, are:

• Descriptive analytics: Can tell you what’s going on in your organization
• Diagnostic: Can tell you why it’s going on
• Predictive: Can tell you what’s going to happen or what will likely happen
• Prescriptive: Can tell you what you should do about it

When inventorying the challenges and needs that an IT analytics solution should address, decision makers can evaluate a solution’s capabilities based on what stage the organization is at, roughly, and where it wants to be in the immediate future. These considerations can allow IT leaders to form criteria that enable an objective comparison between available solutions as well as available analytics vendors and their products.

Descriptive Capabilities — Stage One

The most immediate function an IT business analytics solution can provide is to describe the current state of an organization and its projects descriptively yet succinctly. Baseline reporting functions from IT analytics systems allow leaders to monitor key performance indicators (KPIs) as well as other priority metrics.

These capabilities allow an IT team to answer any number of “what questions,” e.g.:

• What is our incident volume? How has it been changing over time?
• What is the change failure rate for the DevOps teams? Has there been a correlating rise in customer issues?

Without being able to answer these “what” questions, more advanced reporting and predictions produced by the analytics system can lack context. An IT business analytics system at this stage can have relatively advanced reporting capabilities despite lacking more mature analytical potential.

Features and capabilities of a Stage One system can include:

• Integration with key data sources
• Domain-area specific dashboards and reports
• Ad-hoc reporting
• Basic sort and rank functions, such as date/time, online analytical processing (OLAP), and filter expressions

Diagnostic Capabilities — Stage Two

Reporting from Stage One systems can provide descriptive context for an organization trying to monitor the status of its IT functions and projects, but it may leave out critical parts of the picture. For example, the metrics and KPIs being reported on could describe symptoms but fail to reveal the condition or trend making those symptoms appear.

More mature IT business analytics solutions have the ability to dig deeper into data to determine the “why” and more accurately describe trends or past outcomes. This capability grants the opportunity for discovery while also providing a more meaningful representation of data.

Most critically, analytics systems with diagnostic-type capabilities prevent important undercurrent factors in an organization from being overlooked. IT teams who focus just on symptoms could be letting root cause problems linger, creating not just inefficiencies but the risk that underlying problems could simply grow and compound over time.

Mining for root cause of an issue, therefore, is one of the greatest powers a Stage Two system can provide.

A diagnostic-stage IT system can answer “why” questions like:

• Why have our system maintenance costs missed targets for the past three months?
• Why do employees face challenges with sharing and access control functions for certain file types?
• Why do tablet users have more issues with crashing after the recent update but smartphone users do not?

Capabilities defining a Stage Two IT analytics solution can include:

• Ad-hoc analyses
• Data discovery
• Ability to drill down dimensional hierarchy to root level
• Ability to drill across multiple source systems record
• Ability for IT/business users to quickly define own metrics
• Self-service reporting
• Range of visualizations
• Rich interactive geospatial analysis
• Custom data sources
• Prebuilt process mining models
• Near real-time operational reporting

Predictive Capabilities — Stage Three

Stage One IT analytics systems can describe a challenge in hindsight, Stage Two systems can reveal the root cause of the challenge through insight. Stage Three systems provide foresight to prevent issues before they occur thanks to their predictive capabilities.

Being able to predict the performance of future metrics can allow IT teams to anticipate challenges before they occur. In addition, predictive analytics can chart projections for current KPI trends into the future.

Predictive analytics can also demonstrate what might happen in a “what if” scenario, allowing IT leaders to weigh the consequences and benefits of one choice over another. When an accurate prediction cannot be made, some systems can even calculate the approximate level of risk or uncertainty.

One of the most important questions predictive analytics can answer, though, is: “How much risk will X change present? What are the drivers of this risk?” Quantifying change risk using machine learning models can uncover key risk factors held within historical change data and use this information to predict which changes are most likely to fail. With this information, organizations can not only prevent a catastrophic change failure but also monitor for emerging threats to change success.

Specific features and capabilities a Stage Three IT business analytics solution can provide include:

• Prebuilt domain-specific machine learning and AI models
• Metric/KPI forecasting
• Mining of historical data

Prescriptive Capabilities — Stage Four

The term “predictive analytics” once served to illustrate the horizon of machine learning, but organizations quickly learned that making predictions did not always provide the needed information to react to the possible future. To achieve that, machine learning is needed to process potential outcomes and recommend one that provides the optimal mixture of opportunities with minimal risks.

In other words, prescriptive analytics uses predictions based on past data to uncover future opportunities. They generate recommendations for the optimal path to move forward with minimal risk and maximum chances of net positive outcome. On a day-to-day basis, prescriptive capabilities can serve to make IT workflows more efficient while reducing costs.

For instance, by clustering related incidents, IT leaders can learn that some types of incidents have a common root cause and resolution that could be handled by a lower support tier if they were enabled by a knowledge base article. By identifying the “next best action” for incoming incidents, teams can “shift left” to lower level teams, freeing up resources and allowing higher-level teams new opportunities to focus on more important issues.

To achieve true prescriptive capabilities requires more advanced AI models tuned through machine learning and customized to your specific data sources, environment, and market niche. The proper structure of source information can be more important than the specific technology used.

Using IT Business Analytics to Transform from Within

The purpose of ranking analytics capabilities according to maturity stages is to illustrate the transformative potential these systems have.

However, viewing certain capabilities within a strict hierarchy is not necessary, especially since some capabilities could technically represent multiple stages of maturity at once. We say this not to negate the value of Gartner’s analytics maturity model but rather to emphasize that there’s no imperative to prioritize any stage that doesn’t fit within your organization’s specific needs and challenges.

Instead, IT leaders need to focus on the fact that a mature approach to data organization and collection is more important than an analytics system with advanced features. Leaders need the strategy, vision, governance, human skill sets, and technology to support their analytics infrastructure. This effort sets the stage for them to attain their capabilities in a way that achieves the desired effect.

With the right approach and infrastructure in place based on unique priorities, organizations can arrive at the right decision for procuring an IT business analytics solution or modifying their existing one to suit their needs.