What can Data Visualization do to the CPG Industry?

Data Visualization in the era of agents and generative AI

The State of AI in Retail and CPG report of 2024, by Nvidia, talks about how the highest potential for AI impact is in the consumer packaged goods market especially in enhancing the operational efficiency, elevating customer experiences, and driving growth. But it is not just the custom use cases and solutions that AI has improved, the entire Data to Insights lifecycle of CPG along with BI has been impacted.

At the end of the day, we, humans are still visual beings. Even when we’re asking the questions to a generative AI-enabled chat bot, the answer might still end up being a visualization. So, why not understand the nuances to know how to deal with it?

Understanding the supporting pillars of data visualization for CPG

Pillar 1: KPIs of CPG & Retail for analysis

The first step in your visualization journey starts with KPIs. Knowing what you want solves half the problems? (though working on it is the tougher half)

A few KPIs that CPG industry dashboards can have at an operational and analytical level are:

• Product Sales by stores
• Stock Levels for each store
• Product Margins
• Shelf availability
• Delivery time
• Fulfilment statistics
• Customer retention rate
• Preference data (brand vs self)
• Inventory in hand

Some of the KPIs for strategic and financial teams are:

• Sales revenue vs forecast
• Profit per customer
• Product sales by geographies
• Sales order changes (by quarter or month)
• Delivery costs
• Logistics costs to revenue
• Overstock, understock, and deadstock calculations

These are just a few of the KPIs that are useful for the FMCG or the CPG industry. But obviously, these are just an example there might be simpler or complicated insights you might be looking for.

In case, you are not sure about where to start, then one of the best processes would be to identify Acceptable, Advisable, and Admirable KPIs suitable for your CPG business. If you are unsure about the same, drop us a message and our CPG Analytics experts can help you with the same.

Pillar 2: Visualisation representation

Now that you’ve decided on the content or the type of content you want to show, next comes the structure and the placement of the KPIs. Normally, users follow a “Z” pattern in reading therefore the most important KPIs should be placed first and the not-so-important ones at the bottom, like the one in the dashboard below:

Here’s a few tips for visualizing CPG data (in addition to the data and KPIs that have to be represented):

• Lead with the "Big 3" - Start your dashboard with three core metrics every CPG company obsesses over: Distribution (% ACV), Out-of-Stock rate, and Share of Shelf. Put these in big, bold numbers right at the top. If any dip below target (like OOS > 5%), make them turn red.
• Build velocity tiers - Group stores into velocity bands based on sales per square foot (like A: >$750/sqft, B: $500-750/sqft, etc.). Then let users filter all dashboard metrics by these tiers. This helps identify if problems (like out-of-stocks) are happening mainly in high-volume stores.
• Make a planogram health score - Combine metrics like shelf compliance, facing accuracy, and product voids into a single 0-100 score for each planogram. Display this prominently with weekly trends. Regional managers love this for quick store comparisons.
• Include "Never Outs" tracking - Create a dedicated section for your top 20% of SKUs that should never be out of stock. Show their current inventory levels with clear warning thresholds. I've seen companies use a simple red/yellow/green system where yellow triggers if stock falls below 3 days of supply.

Also, it is important to know which type of chart to use when needed. Usage of Pie charts, bar charts, stacked charts, combination charts, waterfall charts, etc. can create confusion when the user is unaware of how to read the charts. Therefore, it is important to maintain charts as simple as possible but also to convey as more information as possible. Here’s a cheat sheet in case you were looking for one.

Pillar 3: Integrations for the Agentic & Generative AI era

The future is going to be a blend of traditional tools with AI, it will be the same with visualizations and agents too. Take our generative AI enabled bot for example:

The best way to get the most out of such agents and bots is, to obviously have the quality data management practices, especially creating and maintaining the semantic models for generative AI purpose. Let’s take 5 examples of in CPG industry:

It is also about embracing the intelligence out there. For example,

• Getting the dashboards scheduled according to the frequency you need them, like weekly & monthly reports. This not only helpful to visualize data in a crisp format but also improves the overall data adoption.
• Voice-enabled data exploration – You can ask natural questions like "Show me which flavors are underperforming in California" and have the visualization automatically adjust. This makes complex data accessible to field teams who may not be data experts.
• Using tools like Power BI to create reports from scratch, based on your inputs and KPIs is very easy (especially with the AI capabilities of Power BI).

The examples are not just limited to visualizations, you can use AI agents for automatic anomaly detection – instead of waiting for humans, you can trigger agents for early detection of anomalies, and more.

How we make this happen?

We understand that all this might sound very tedious, right from enabling the data management to creating the supporting CPG applications that enable decision making. So, we’ve introduced 1Platform which is an ecosystem that supports CPG Data visualization and AI seamlessly.

The short version of it is you can access everything you need right from AI enabled applications to AI enabled dashboards from the same place.

CPG Industry Dashboards: BI Portal in 1platform

With 1platform we eliminate the need for searching multiple reports but instead our navigation-led Insight Portal, we give you a very clear demarcation about what you should be looking for.

It’s similar to having separate tabs for visualizations like: Shipment vs Depletion or a Sales overview tab.

You can drill down further based on your requirements.

Will dashboards be dead?

Or will they be so embedded into our workflows that we barely notice it anymore? Or some mixed version of it? We’ll have to see.

But in order to adapt to the AI era, you need to move quickly!!!

Talk to our AI & BI experts today to discuss your requirement further!!