Reporting & BI tools

Most BI tools provide a graphical interface on top of an existing database or data warehouse, allowing users to create reports and dashboards without directly writing SQL queries.

They are commonly used to monitor Key Performance Indicators (KPIs) such as:

• Sales evolution over time
• New customers acquired each month
• Customers lost each month
• Churn rate
• Other operational and business KPIs

Behind these visualizations, the calculations are typically performed by the underlying data warehouse. To avoid repeatedly running resource-intensive queries, frequently used results can be precomputed and stored in dedicated tables that are periodically updated.

Reporting tools vs. BI tools

Traditional reporting tools generally produce static outputs such as Word or Excel reports. BI tools add an interactive layer, allowing users to explore data directly through dashboards.

For example, users can click on a location, category or metric to progressively access more detailed information. This type of interaction is commonly referred to as drill-down.

Some BI tools store the data being analyzed directly in RAM instead of querying a database. This approach is generally referred to as In-Memory Analytics.

It can provide faster dashboard calculations when datasets are small enough to fit into memory, but it also introduces important limitations:

• Large enterprise datasets may exceed the available memory.
• Data may need to be duplicated between the central database and the BI platform.
• Maintaining synchronization between multiple copies of the data can become more complex.
• The performance advantage can be less relevant when standard reports are refreshed periodically rather than calculated interactively.

For more advanced use cases such as customer segmentation, cross-selling, churn prevention or up-selling, TIMi provides dedicated analytics capabilities designed to explore large datasets and identify relevant segments.

Terminology note

BI tools are sometimes referred to as analytical tools. This can create confusion with platforms used for predictive analytics, modeling and scoring. Throughout TIMi Academy, we distinguish Business Intelligence and reporting from Advanced Analytics and Predictive Analytics.

The differences between different BI tools are:

• Is it “In-RAM-BI”?
  If yes: limited scalability, marginal speed increase
  If no: unlimited scalability (depends on the underlying database)
• Charts: Some BI software propose only a small sets of charts
• Report Exportation:Some BI tools are not able to export their reports into standard formats like: Word documents, Excel spreadsheets, Powerpoint presentations, HTML pages,…You must check carefully your BI tool to see if it’s able to export its report to Microsoft Office documents because most of them are not able to do that. It’s important to have reports as Microsoft Office documents to be able to work collaboratively on the report with your colleagues (e.g. adding annotations and notes that other can see). …or to publish it on a sharepoint portal, for example.

A basic BI tool is included inside Anatella (that is included inside the TIMi framework). With Anatella, you can create any chart that is inside any Microsoft Office document (Word, Excel & Powerpoint). Anatella is not an “In-Ram-BI” tool and thus it’s not adapted to do exploratory analysis of our database. Anatella is scalable and can analyze any size of database (as all ETL’s) to produce the required reports. Hopefully, the TIMi suite contains two other tools specialized in exploratory analysis of your data: TIMi and Stardust. Typically, Anatella is used to automatically refresh every day or week standard Powerpoint reports that give detailed information about the evolution of different KPI’s.

The differences in results between all these BI tools are:

None: Given enough time and enough human resources, you can create similar reports with all the above tools. The speed of the tool mainly depends on the underlying database. So, I would suggest you to invest in a good database solution rather in an expensive BI tool. Even for limited In-Memory BI tools, there exists some techniques to get around the size limitation of these tools.

If you use the same source data, you’ll obtain the same charts, whatever the BI tool used. This seems logical. The same is not true about Advanced Analytic Software: When working on the same dataset and using the same modeling methodology, you’ll still obtain different predictive models, with different ROI, depending of the modeling tool used. This difference can be very significative: On most cases, TIMi give 5 to 15% more ROI.

One important thing that very few people know is that there is a good BI tool that is integrated inside Microsoft Excel! Microsoft Excel is able to use your database “in direct” without any exportation, like any other classical BI tool would do. With Microsoft Excel, you can directly connect to your database system and easily design with your mouse very complex BI charts.

Although the differences between the different BI tools are inexistant in terms of end-result, you can still observe large differences when it comes to: speed, cost of ownership, quality of the interface (“user-friendliness”), energy efficiency, beauty of the generated reports (is it eye-candy?), etc.