The Future of EPR Analytics: Why Companies Need Packaging Intelligence, Not More Spreadsheets

Extended Producer Responsibility (EPR) is quickly becoming more than a regulatory requirement. It is becoming a significant data-management challenge.

As more states implement EPR programs for packaging and paper products, companies must collect, organize, validate, and report increasingly detailed information about the packaging they place into the market.

For producers selling hundreds or thousands of SKUs (Store Keeping Unit), this creates an important question:

How many separate EPR spreadsheets can a company realistically maintain?

I believe companies need to stop thinking about EPR reporting as a collection of spreadsheets and start thinking about it as a data system.

And AI is going to play an important role in building that system.

The Real Challenge Is the Data

On the surface, EPR reporting can appear straightforward.

Determine how much packaging a company sells into a state, classify the packaging by material, and report the required information.

The reality is much more complicated.

The information needed to answer those questions may be scattered across multiple systems and departments.

A company’s financial database may contain SKU numbers, UPCs (Universal Product Code), descriptions, sales units, and revenue.

Packaging specifications may exist somewhere else.

Component weights may be stored in technical documents or supplier specifications.

PCR (post-consumer recycled) content may come from another source.

Information about recyclability, packaging construction, coatings, labels, sleeves, closures, cartons, blisters, and other components may not exist in a structured database at all.

And the identifiers used by one system may not perfectly match the identifiers used by another.

Before a company can perform EPR analytics, someone has to connect all of this information.

That is often the hardest part.

I’ve Seen This Problem Firsthand

At Topline Statistics LLC, I have worked directly with EPR packaging data and built databases containing thousands of SKUs.

That experience has shown me that EPR analytics is rarely a matter of receiving one clean dataset and calculating a few totals.

It is much more likely to involve multiple datasets that were never designed to work together.

One source may identify a product by SKU. Another may use a UPC. Packaging information may use a different part number. Sales data may have abbreviated descriptions. Some packaging components may have detailed specifications, while others have incomplete or missing information.

The first challenge is therefore not statistical analysis.

It is building a reliable database from fragmented information.

That is the same basic problem I solve throughout my work at Topline Statistics: organizing raw and disconnected data into a structure that can actually be analyzed.

One Product Can Become Many Data Records

A single SKU might contain:

• A plastic bottle
• A plastic cap
• A label
• A shrink sleeve
• A paperboard retail carton
• An instruction insert

Each component can have its own material, weight, PCR percentage, recyclability characteristics, and EPR classification.

Now multiply that by hundreds or thousands of SKUs.

Then consider multiple states with different requirements, reporting categories, fee structures, and implementation schedules.

The complexity grows very quickly.

This is why I believe the future of EPR compliance will depend heavily on good database architecture.

Build One Packaging Intelligence System

Rather than creating another spreadsheet every time a state implements EPR, companies should work toward creating a centralized packaging intelligence database.

At the center of that database should be the product.

Each SKU can then be connected to its individual packaging components and their attributes.

For example:

Product → Packaging Component → Material → Weight → PCR Content → Recyclability → State Classification → Sales Volume → EPR Fee

Once that structure exists, new EPR requirements become much easier to manage.

When another state implements an EPR program, the company should not have to rebuild its packaging data from scratch.

Instead, analysts can determine which additional classifications, calculations, or reporting fields are required and add them to the existing system.

That is a much more scalable approach.

This Is Where AI Becomes Powerful

This is also where my approach to EPR database development has changed.

AI can dramatically accelerate the process.

Consider a company with several thousand products and packaging components. Instead of manually researching and comparing every record, AI can assist with:

• Matching SKUs across datasets when product numbers or descriptions do not match perfectly
• Identifying missing packaging information
• Researching packaging materials and components
• Standardizing product descriptions
• Flagging inconsistent classifications
• Detecting unusual component weights
• Building formulas and calculations
• Restructuring datasets
• Comparing regulatory requirements
• Identifying records requiring human review

Tasks that once required hours of manual searching and comparison can increasingly be completed much faster.

But there is an important distinction.

AI is helping me build the database. It isn’t making the final decisions for me.

AI Still Needs Someone Who Understands the Data

Consider some seemingly simple packaging questions.

Is a bottle PET or HDPE?

Does it have an unmarked plastic blister PET, or should it be classified more conservatively?

Is a shrink sleeve part of the bottle or a separate packaging component?

Does the stated PCR percentage apply to the entire package or only one component?

Which sales dataset represents the units actually sold into a particular state?

Those decisions can materially change the final EPR calculation.

AI can investigate the question, identify patterns, compare records, and recommend an answer.

But an experienced person still needs to establish the classification rules, challenge questionable results, document assumptions, and validate the final database.

That is why I don’t see the future as AI replacing EPR analysts.

I see a small number of knowledgeable people using AI to manage an amount of data that previously would have required a much larger team.

From EPR Compliance to EPR Analytics

This is where I believe companies have an opportunity to get much more value from the work they are already required to perform.

Once packaging components, materials, weights, PCR content, recyclability, sales volumes, and EPR fees are connected, the database becomes more than a compliance tool.

It becomes an analytical tool.

Companies can begin asking:

What products generate the highest EPR fees?

Which materials account for the greatest percentage of our packaging weight?

Which packaging components are hurting recyclability?

Where would increasing PCR content have the greatest impact?

Which SKUs contain unnecessary packaging?

Which packaging redesign projects could potentially reduce EPR fees?

That is the difference between EPR reporting and EPR analytics.

Reporting tells you what happened.

Analytics helps you decide what to do next.

How Topline Statistics Approaches EPR Data

At Topline Statistics, my services are built around three areas: Data Organization, Data Analysis, and Data Reporting.

EPR requires all three.

Data Organization

Packaging specifications, SKU lists, sales data, material classifications, component weights, PCR information, and regulatory requirements must first be connected into a usable database.

Data Analysis

Once the database exists, the data can be used to calculate packaging weights, PCR percentages, recyclability metrics, state volumes, EPR fees, and opportunities for packaging improvement.

Data Reporting

Finally, thousands of individual records need to be transformed into information that sustainability teams, packaging engineers, managers, and executives can actually understand and use.

AI can accelerate each stage.

But the quality of the final analysis still depends on the quality of the underlying database and the judgment of the person building it.

The Database Can Become More Valuable Than the Report

This may ultimately be the biggest opportunity created by EPR.

Companies are being required to collect packaging information that many organizations have never connected before.

Once that information has been assembled correctly, why use it only once for compliance?

A well-designed EPR database can become a long-term packaging intelligence system.

It can support regulatory reporting today while helping companies make better packaging decisions tomorrow.

And as additional states implement EPR programs, that centralized database becomes increasingly valuable.

The Future Is Human + AI

I don’t believe companies need an army of people manually maintaining EPR spreadsheets.

I also don’t believe they can hand their packaging data to AI and expect an accurate compliance database to appear.

The better model is somewhere in between.

A small number of people who understand packaging, regulations, data, and business operations can use AI to perform much of the repetitive work involved in building and maintaining these systems.

AI provides the speed.

Humans provide the judgment.

And a well-designed database provides the foundation.

This is the direction I see EPR analytics heading, and it is the type of work I am continuing to develop through Topline Statistics LLC.

As EPR expands across the United States, the companies that get ahead will not necessarily be those with the largest compliance departments.

They will be the companies that figure out how to turn fragmented packaging information into reliable, reusable data.

The future of EPR analytics isn’t more spreadsheets. It’s packaging intelligence.

Need Help With EPR Data?

Topline Statistics helps businesses organize complex datasets, analyze the information behind them, and turn the results into clear, actionable reporting.

My experience includes building large EPR packaging databases, connecting fragmented product and packaging information, analyzing packaging fees, and developing reporting systems that transform thousands of individual records into useful business insights.

Learn more about Topline Statistics LLC and my data consulting services at toplinestatistics.com.