Analytics Maturity Canvas

How Do I Use the Analytics Maturity Canvas?

The Analytics Maturity model helps you to find out:

1. Which analytical maturity does my company, do individual departments of my organization or my customers have?
2. Which are possible data-driven applications and their requirements for the company, the department or the customer regarding the analytical maturity level?
3. Which are the relevant use cases to take my company, departments or customers to the next analytical level?
4. Which analytical tools do I need for the realization of the applications?

For more information, see Data Strategy Design.

Start

Start working on the Analytics Maturity template by documenting the current state of your company, department, or customer. Focus your attention on the following two aspects:

1. Use cases: which analytics solutions or applications are already in use in your company and for what purposes? To do this, fill the boxes (application fields) in the upper green area (=utilization) from left to right with the appropriate cards.
2. Tools: which analytical tools, procedures, or software systems do you use for this? You answer this question in the lower yellow area (=refinement).

The boxes reflect the level of analytical maturity, and the analytical maturity level determines the complexity of the particular application or tool. The analytical maturity levels (yellow area below) are interdependent: to make predictive analytics (predictions), you first need diagnostic tools to identify patterns in the data. Correspondingly, companies (green area above) also undergo a maturing process (from left to right). As the complexity increases from bottom to top (yellow area), the value-added from the data (green area) increases from left to right. It is the goal of companies to continuously increase their maturity level by introducing appropriate tools and applications based on them. A detailed explanation of the individual maturity levels and maturity process steps can be found below.

After you have documented the actual state, start by specifying the target state by filling the green boxes at the top and the yellow boxes at the bottom – from left to right or from bottom to top. Use different colors for the cards:

• Green: existing applications or tools
• Yellow: applications or tools which are planned or whose development is in progress
• Red: required or desired applications or tools

Pay attention to consistency while filling it in:

1. There must be an appropriate analytics tool for every analytics application. For example, for a marketing KPI dashboard, you might need a (self-service) Business Intelligence tool.
2. Conversely, there should be at least one analytical use case for an analytical tool. Otherwise, this means that the tool is not in use and therefore may produce unnecessary costs for your business.
3. The maturity level of the analytical application must match the maturity level of the analytical tool. To make predictions, such as revenue forecasting, you need a predictive analytics tool.

Step 1 of 11

Business Applications

The company runs business applications without analyzing the data created as part of its business operations. Examples of business applications include CRM and ERP systems, corporate websites and apps, IT-controlled production systems, etc.

Examples of applications:

• CRM software
• ERP systems
• Corporate websites
• E-commerce websites
• Mobile apps
• IT supported production systems

Step 2 of 11

Descriptive Analytics

Descriptive analyses describe what has happened. They deal with metrics and key performance indicators (KPI) which measure the progress towards objectives. Descriptive analysis provides a retrospective view of what has happened within the company and in the market. Descriptive analytics typically uses anonymized and aggregated data. The analysis’ results still require a high degree of interpretation by the decision-maker. Often a descriptive analysis is followed by a diagnostic analysis, for example, to investigate the reasons for having failed to reach an objective. Descriptive analyses are often based on metric frameworks such as (value) driver trees, balanced scorecards, or the AARRR model (see Customer Touchpoints Canvas).

Examples of tools:

• Classical Business Intelligence solutions
• Manually created reports in spreadsheets
• (Online-) Analytics services
• Reporting tools

Step 3 of 11

Business Reporting

The company calculates metrics and key performance indicators (KPI) based on the data coming from the business applications as well as from other sources. These are then, for example, included in the weekly or monthly business reporting (classic Business Intelligence).

Examples of applications:

• Marketing KPI dashboards
• Sales reports
• Production statistics

Step 4 of 11

Diagnostic Anayltics

Diagnostic analyses explain why something has happened. They deal with patterns in the data such as trends, correlations, outliers, etc. Diagnostic analyses provide insights into the mechanisms of a company and a market. Diagnostic analytics typically uses non-aggregated data. A decision-maker uses the results of diagnostic analysis to plan and adapt future measures. In order to decide which metrics are worth a diagnostic analysis, descriptive analyses often precede the diagnostic analysis.

Examples of tools:

• Business discovery applications
• Self-service BI solutions
• Statistical tools
• Testing tools
• Visualization tools
• Analytical programming languages

Step 5 of 11

Business Discovery

The company analyzes the metrics and key performance indicators (KPIs) looking for anomalies and explanations helping to improve its business processes and business model. In order to do so, employees are using, for example, interactive dashboard applications.

Examples of applications:

• Ad hoc analysis for customer insights
• A/B testing of website designs
• Multivariate testing in online advertising
• Correlation and trend analysis based on social media monitoring
• Analysis of causes of defects in production
• Customer segmentation

Step 6 of 11

Predictive Analytics

Predictive analyses predict what most probably will or could happen. They create statistical or stochastic models to predict values and their probability. A predictive analysis provides an outlook on future developments. Predictive analytics is based on non-aggregated and often non-anonymized data. A decision-maker evaluates options for actions based on the predictions and decides and/or acts accordingly. For modeling, diagnostic analyses often precede a predictive analysis in order to identify patterns and verify them with domain and/or industry know-how. In production, descriptive analyses validate predictions and check their utility.

Examples of tools:

• Predictive modeling tools
• Data mining tools
• Data Science suites
• Analytical programming languages
• Machine-Learning-as-a-Service

Step 7 of 11

Business Forecasting

The company uses its data to predict future company and market developments. To do so it uses advanced analytics techniques (business analytics). Examples are sales forecasting, churn prediction, lead scoring, or predictive maintenance.

Examples for applications:

• Sales forecasts
• Prediction of churns
• Customer rating (“lead scoring”)
• Calculation of the customer lifetime value

Step 8 of 11

Prescriptive Analytics

Prescriptive analyses recommend what should happen. They evaluate options based on predictions (predictive analytics), simulate different scenarios, and provide recommendations for actions based on the simulated results. Prescriptive analyses still require a decision-maker to make the decision regarding an option as well as to take the subsequent action.

Examples of tools:

• Modeling tools
• Simulation tools
• Data Science suites
• Analytical programming languages

Step 9 of 11

Business Optimization

The company looks for optimization potentials within the data. To do so it simulates potential measures and/or solutions and analyzes the outcomes. The measures and/or solutions with the most favorable outcomes, according to the simulations, are subsequently implemented.

Examples of applications:

• Controlling of advertisements
• Recommendation systems for products
• Predictive maintenance
• Churn prevention

Step 10 of 11

Automated Analytics

Autonomous analytics decides what should happen. It uses prescriptive analytics and carries out the action by itself. A decision-maker is not involved.

Examples of tools:

• Marketing automation solutions
• Service orchestration tools
• Agent systems
• Artificial Intelligence systems

Step 11 of 11

Business Automation

Machines control processes and their optimization autonomously. The company’s employees have only a monitoring function.

Examples for applications:

• Dynamic pricing
• Autonomous vehicles or aircrafts
• Intelligent storage
• High frequency stock trading

Completion

After you have completely documented the actual status and defined the target state sufficiently, you should once again carry out a final consistency check (see checklist above). The next crucial step is to select the applications which will increase the analytical maturity of your business – without skipping a step to minimize risk, effort, cost, and project duration, resulting in a rapid return on investment, and to benefit from the analytical optimization potential in a timely manner.

Rather take many small and fast steps versus one big and slow step. Anyone who has ever tried to take two or more steps at a time on stairs knows how exhausting it is and how dangerous it can be.

Therefore, when selecting the relevant applications, the following points should be considered:

1. Use cases in applications with a high density of green cards (i.e. existing applications and solutions) are likely to only marginally increase the company’s efficiency and effectiveness. The potential for optimization, for example through Business Intelligence solutions, is already exhausted if a corresponding reporting system has already been introduced for all departments.
2. Unoccupied fields of applications or boxes with many red cards (i.e. missing applications), however, suggest a high value creation potential.

3. However, it must be checked whether the corresponding analytical tools for this purpose are available in the company. For example, optimization applications require prescriptive analytics tools. If these tools are not yet available, they must first be bought and implemented in the company. This usually requires a costly evaluation of the tools as well as time consuming training of employees.
4. Accordingly, no analytical maturity level should be skipped. For example, if the company uses only descriptive analytics tools for reporting, and wants to use predictions, i.e., develop predictive analytics solutions, to take two levels in the maturity model, then first introduce diagnostic tools and identify use cases to directly benefit of these tools and to gradually increase the degree of maturity in the company and in particular among the employees.

Finally, you select the use cases that take the maturity of the company to the next level. You can then use the Data Strategy canvas to specify the respective use cases.

References:

• Value Proposition: for example, to identify data-driven applications and solutions (value propositions) for customers offered as products or services.
• Business Model: key activities such as marketing controlling require analytical applications (for example, KPI reporting). Customer relationships can be also based on analytical applications such as chatbots. After all, the analytic solutions may be a key resource.
• Data Strategy: to specify specific use cases and to determine the required analytic maturity level (i.e. the analytics tools).
• Data Landscape: to identify possible use cases based on the available data.