Narrative Enabling Data Visualization | Maavrus

Written by Mahadevann Iyerr

April 7, 2022

Data visualization

No one ever, made a decision because of a number. They need a Story – Daniel Kahneman.

Most business outcomes are the result of a combination of multiple-input factors. Each input is in turn performed by various people & systems. The analysis of any business performance is based on various hypotheses, learned from practical business experience. A narrative building data visualization approach tries to validate or nullify the hypothesis, by leveraging visuals and graphs and shines the torch on the most important factors

This is easier said than done. Apart from business outcome and input factors, one needs to also consider the various stakeholders who will consume the analysis. Also, in many cases the causation may not be clearly attributable, so realistic and clearly articulated assumptions can help substantiate the narrative.

I will try and explain this using a business scenario, in which I was personally involved.

In retail grocery hypermarkets, price integrity issues are a known challenge. Essentially price integrity issue means that there is a difference between the price mentioned on the shelf for that product and the price that is charged at check-out. If the price at check-out is higher it is both a regulatory non-compliance as well as a reason for customer angst.

With most hypermarkets, having over 50K SKUs and hundreds of stores, and daily price changes, updating price labels is a herculean task. These errors can happen, because of human error when manually changing shelf edge price labels, delay in data visualization sync issues between the price file and the label printing and PoS systems, OR erroneous data entry on the label printing systems. Every time a customer notices this difference, she would bring it to the cashier’s notice, who in turn would refund the difference to the customer, and update it in the system as a price integrity ticket

A few years back there was a huge uproar in the media around price integrity in Grocery formats, which led to initiatives being undertaken on a war-footing both in stores as well as at central functions. While manually updating shelf edge labels was in store control, most other reasons were centrally controlled by the commercial and IT teams. With the level of leadership focus, the number of price integrity tickets per day started to reduce week on week, but the challenge was that one did not know, which of the initiatives had started to deliver. The root cause analysis of each of the tickets was an involved and time-consuming affair, and we needed to quickly understand if the tickets per day were reduced on account of operational excellence in stores or any centrally driven initiative.

Table of Contents

Data visualization

Every single day, we bucketed the SKUs with price integrity issues, as SKUs that had price integrity issues in a single store, in 2 stores, and 3+ stores. In trying to visually depict this for the leadership team, we made certain assumptions. In the SKUs where the price integrity issue was there in only a single store, we assumed it was predominantly a store execution issue, and the SKUs where we found the price integrity issue for that SKU in 2+ stores we assumed that it was on account of a central issue; since the probability that the same shelf edge label, not being manually updated across multiple stores was fairly remote.


Numbers are fictional and used for the purposes of explaining the approach

With this data visualization, we were able to quickly conclude that the execution excellence in stores was the main reason for the reduction in price integrity issues, and the central initiatives were yet to make a meaningful contribution in the first 30 days. The trend line for “number of SKUs with a single store price integrity occurrence”, was also used as a measure to continue tracking the rigor of shelf label execution in stores.

Hope this example helped articulate, how one can combine the end objective of the analysis and realistic assumptions, in identifying patterns quickly and enabling agile business decisions.

Related Articles

Customer feedback analytics to influence D2C product development​

Customer feedback analytics to influence D2C product development​

Hi. Welcome to expert talks at Maavrus, in this video, we’ll talk about how Direct to Consumer or D2C businesses can better understand their customer feedback data and leverage it to develop their existing and new products. https://www.youtube.com/watch?v=bd4PqAKjfTQ...

Analytics to grow Digital Commerce and Improve Digital Marketing ROI

Analytics to grow Digital Commerce and Improve Digital Marketing ROI

Hi. Welcome to Expert Talks at Maavrus, in the previous video, we spoke about why business leaders and marketing teams will need to have a mindset shift in the way they look at the effectiveness of their marketing spends. https://www.youtube.com/watch?v=SfwtRtjaQWk...


Submit a Comment

Your email address will not be published. Required fields are marked *