8 best practices for optimizing your analytics reports

Are your reports delivering the right information for your business? Here's how to get the best information into the correct hands.

Digital marketing campaign data analytics report with metrics and key performance indicators (KPI) on information dashboard for advertisement strategy on internet, business person in office

Image: iStockphoto/NicoElNino

While we work hard to produce important and useful data analytics reports, we know that not all of the data we present is being used to its fullest extent. Three years ago, Gartner reported that up to 97% of unstructured big data was not being used in enterprises. This number has gone down since, but now the question is: How much of the data that we analyze and report on is being actively used?

The report usage question has plagued IT from the beginning. Most IT staff know that the 80:20 rule applies: 20% of the reports produced for the business do 80% of the informing. Meanwhile, unused and seldom-used reports pile up on servers.

SEE: Report: SMB's unprepared to tackle data privacy (TechRepublic Premium)

There is no reason to believe that big data reporting is any different. To prevent wasted efforts in dashboard and analytics report development, here are eight best practices to follow:

1. Stay tuned with the business

How many times does IT meet with users about a report design, and then go away to develop something else? More often than you think.

What happens is that IT, as it works on the report back in the office, thinks of new ways to slice and dice the data and decides to embellish the original request with additional functions and features. 

This is a great practice—and can pay off "big" for users—as long as the embellishments don't create so much report drift that the original business request is missed.

SEE: 7 big data wishes for 2021: IoT standardization, stronger use cases, and more (TechRepublic)

2. Visualize dashboards and enable easy drill-down

Financial departments are comfortable working with spreadsheets and figures, while sales might prefer a pie chart, manufacturing might prefer bar charts, and logistics might prefer a worldwide map.

Finding the optimal visualization of summary level data for each user is major victory in itself. It immediately creates a level of comfort for the user.

Another usability factor is an easy drill-down into more detailed analytics data. For example, if a user is working with a map summary visualization and wants to know more about his truck fleet in Atlanta, he should be able to click on Atlanta so he can get to the detail.

3. Ask next-generation report questions

Today, your user might be asking for a report that tells him how much product flows through each of his production lines hourly, daily, and monthly. Next year, he might want to know how much product was returned for defects and which production lines produced it.

From a data standpoint, and also from a report data field definition standpoint, it's always a good idea to ask the user what he might want to see from a given report in the future so you can easily scale to that and keep the report relevant.

4. Enable multi-level usage clearance and universal access

At any given point, a new user in a new business area might request access to a report. At all points in time, the controlling user of a given report will also want to give out security clearances at different levels to people, such as a VP of manufacturing being able to see all manufacturing activity, but the manager of Plant B only being able to see information for Plant B.

Analytics report designs should clearly designate security access levels, and who should control and authorize them. These reports should also have the technical flexibility to be accessed by anyone in the enterprise who is cleared for use. 

SEE: Data analytics' big problem: 'The tools are nice, but how do you get people to use them?' (ZDNet)

5. Verify data integrity

Before any analytics report or dashboard is cleared for use and moved to production, the data that it uses and reports should be cleaned and verified for accuracy.

6. Synchronize data with like data in the enterprise

If your sales reports use the data field of "customer," which refers to individual buyers, and your manufacturing systems uses the term "customer," which refers to individual buyers, but also to a rework shop within the company, this data should be synchronized so there is a common definition that enable sales and manufacturing to talk about the same thing. 

Data synchronization is done in the database area of IT. It's important because information discrepancies and internal disagreements can arise when two different departments think they are talking about the same thing but they aren't. 

SEE: Your big data reports are not being used as they should: Here's how to change that (TechRepublic)

7. Standardize report development and formats

Standardize the report-producing tools that you use and also the formats that various reports use. This helps ensure uniformity across the enterprise, and lessens confusion for users.

8. Measure for use and perform post-mortems

Annually, IT should review analytics reports for the amount of use they're getting. If a report hasn't been used, or was seldom used, IT should check with end users to see if the report is still relevant.

It's equally valuable to conduct a post-mortem evaluation. Which content, feature, and function characteristics of the reports were most widely used? What did you learn about the reports that weren't used? What can you take away from the evaluation to improve the quality of analytics reports? These are all important questions to ask to ensure the reports meet the end users' needs.

Also see