Chapter 4. Are You Ready for Converged Analytics?
In Chapter 1, you learned how organizations across different industries assessed their current analytics strategies to take the right next steps. An analytics maturity model can be used to situate and track this progress. One such model comes from Thomas Davenport and Jeanne Harris’s book Competing on Analytics: The New Science of Winning (Updated, with a New Introduction).1 This model identifies five stages of analytical competition.
Earlier chapters described analytics strategies as a combination of people, processes, tools, and data. You will see the same factors at play in reaching each stage of analytical competition. As organizations progress through the stages in Figure 4-1, they tend to adopt more advanced analytics.
The lowest stage of analytics maturity is being analytically impaired. In this stage, analytics is left to the experts, who themselves only receive the data after lengthy requirements gathering processes, which are hard to modify once implemented. Any resulting data products tend to be siloed, with ad hoc reports and analyses based on purely historic data.
Figure 4-1. Five stages of analytical competition (Adapted from Davenport and Harris, 2017)
The next stage of analytics maturity, localized analytics, offers somewhat more reliable processes for collecting and reporting on data, but these efforts are still isolated. These earlier stages of analytics maturity resemble what the organizations might face before adopting a converged analytics.
When an organization stands at the analytical aspirations stage of analytics maturity, it’s adopted the commitment to a broader use of data but doesn’t necessarily have the combination of people, processes, tools, and data to support it. An organization at this level of maturity may have a proliferation of data sources and software that data scientists and business users interact with independently. At this stage, an organization may have adopted a more advanced data warehouse or have struggled to successfully implement a data lake.
As an organization matures to become an analytical company, it adopts an enterprise-wide priority for integrated and governed data. It begins to offer embedded analytics products to users under a converged architecture. Finally, an analytical competitor routinely benefits from its organization-wide analytics efforts. It has democratized how data and analytics are used. Importantly, an analytical competitor only sustains that advantage with continuous improvement through its people, processes, tools, and data. Backed by data, the organization can adapt and respond to a changing business environment.
Six Dimensions for Benchmarking Maturity
To benchmark where your organization fits on Davenport’s analytical competition model, six dimensions are described here. Table 4-1 then describes how to assess an organization on those dimensions for different levels of Davenport’s model.
- Culture and organization
-
Not everyone in an organization needs to be trained as a data scientist to get massive value from data. As an organization matures, it gives its “citizen data scientists” resources to lead with data.
- Key challenges
-
Simply gathering the data to build a model is difficult in less mature organizations. As an organization matures, it’s able not only to deploy models but to update and audit them: a critical ability when business conditions change rapidly.
- Information availability
-
An expanding source of data for analytics should be a blessing, but without consistent data governance can be a curse. A mature organization doesn’t just make data available, it has policies to define and protect the data.
- Complete understanding
-
In less mature phases of analytics, users can only monitor the business through the rearview mirror of static, backward-facing reports. As the organization matures, users benefit from streaming data and embedded AI to view the business as it happens.
- Increased business agility
-
This report has emphasized how an organization can become more agile through the adoption of advanced analytics. With the right combination of people, processes, tools, and data, organizations are able to, in the words of Heisterberg and Verma, “anticipate challenges and opportunities before they occur.”
- Materialized value
-
Collecting more data is one thing; a truly mature analytics organization shows tangible output from their efforts. Getting tangible, materialized value is ultimately the goal for a converged approach to analytics.
Based on Table 4-1, where does your organization fall along these six dimensions?
| Analytically impaired | Localized analytics | Analytical aspirations | Analytical company | Analytical competitor | |
|---|---|---|---|---|---|
Culture and organization |
Only technical experts can lead with data |
←→ |
Business users have some data availability, and their work is assisted by data scientists or IT |
←→ |
“Citizen data scientists” can operate autonomously with a variety of data products |
Key challenges |
Difficulties gathering data to use in models |
←→ |
Difficulties putting models into production |
←→ |
Difficulties auditing and continuously updating models |
Information availability |
Data is siloed and not uniformly available |
←→ |
Data is widely available but without consistent governance |
←→ |
Data is well-governed, integrated, and available |
Complete understanding |
Business users rely on static, backward-facing automated reports |
←→ |
Business users can interact and ask questions of historical data |
←→ |
Business users have complete view and command over historical and real-time data |
Increased business agility |
Static reporting tools provide limited view of business conditions |
←→ |
Dashboards and BI allow for more fluid understanding of business conditions |
←→ |
Visual analytics, embedded AI, and real-time analytics increase business agility |
Materialized value |
Organization is slow to gather value from data-backed insights |
←→ |
Business users rely on limited, primarily backward-facing data |
←→ |
Models are quickly deployed and blended with reporting and BI to provide an integrated view of the business |
Establishing a Framework for Analytics Maturity
Earlier chapters of this report view past, present, and future analytics through a lens of people, processes, tools, and data. This chapter explained how as an organization adopts more modern analytics, such as converged, it becomes more mature.
But how exactly do organizations align people, processes, tools, and data to level up their use of analytics? It’s easy to place primary emphasis on raw data and technology—as more data became available, the organizations that implemented the right systems to collect that data became mature. But the ability to collect data is not the same as the ability to use it. In fact, as data grows in variety, volume, and velocity, more emphasis must be placed on governing and auditing its uses—only then can it be used reliably to drive business insights.
No analytics transformation can happen without a growth in data and the tools needed to collect the data. But even more important are processes for working with data that are transparent and governed to empower individuals of every role and skill level to lead with data. By reducing the use of “black box” algorithms, and by increasing explainability and transparency, organizations can reap the benefits of AI while reducing bias risks.
Chapter 5 provides first steps toward making these alignments.2
1 Thomas Davenport and Jeanne Harris, Competing on Analytics: The New Science of Winning: Updated, with a New Introduction. (Boston: Harvard Business Press, 2017).
2 TIBCO has provided an Analytics Maturity Assessment to identify areas for improvement in your analytics strategy and benchmark it to others in your industry, with an emphasis on unlocking agility.