I lied; it’s actually six questions. The first one is simple though: 

What is Master Data Governance?

Master Data Governance: The framework of policies, processes, and standards that ensures high data quality, security, and compliance within an organization.

Talking with brands about master data governance when it comes to their product catalog tends to elicit one of two feelings: fear, or boredom. 

The fearful response comes from the database managers, or those involved day-to-day with product enrichment; people who understand the unpleasant truth of “how the sausage is made.” These individuals often aim to go all in, wanting to cleanse and standardize every piece of product data. They’re scared because they know just how much work this requires. 

In contrast, the bored response arises from business users or leadership who are more detached from the realities of the data and primarily concerned with an impending Go-Live date. These folks often prefer to ignore data governance altogether, or pretend that AI can do it all for them (hint: it can’t). Neither of these differing approaches is realistic for most organizations. A middle ground must be found. 

We could write a book about what a proper master data governance policy entails (people have). You need to account for things like data ownership, lifecycle management, auditing and compliance, collaboration policies, integration standards, training procedures and more. 

But many brands don’t have the time or resources to build out a full-fledged data governance policy for product data. So how should you incorporate data governance into your PX Strategy? 

Below is the cheat sheet version of data governance for brands trying to move quickly and efficiently. By addressing these questions in your product catalog and implementing ongoing controls, you’ll be able to realize the most important benefits of data governance without incurring the time or expense of implementing a full-fledged program for your Product Data.  

So, ask yourself….

1. How will you ensure your product data is complete?

One of the quickest ways to harm your reputation with a consumer is to present them with an empty field instead of the product information they expect. A missing image, product name, or critical technical spec can crush your conversion rate. It’s critical that you have a mechanism in place that will prevent you from sending incomplete product data to customers.

 Many brands make the unfortunate mistake of creating large “Description” text fields that are ultimately nothing more than a jumble of ungoverned data. They have no way of knowing what is or isn’t there. For proper governance, these key product data points should be separated into discrete fields, allowing easy validation of their completeness. A robust Product Taxonomy with appropriate attributes and differentiation between product types is essential to help ensure this completeness across an entire catalog. 

Once individual fields are validated, they can be concatenated programmatically to build descriptions, or even serve as prompts for a generative AI solution.

 2. How will you keep your product data consistent?

Your product data must be consistent AND displayed consistently. Product data goes to many different endpoints within an ecosystem, and without a PIM system providing a single source of truth, it’s not uncommon for siloed teams to create different variations of product data for the same product. This not only creates inefficiencies, but also inconsistent experiences that can deter customers.

Just as critically, you must ensure that data itself is consistent. In the governance process, each data field format must be identified and assigned a standard that is always enforced. For instance, a data field like ‘Seat Height’ for a furniture distributor should convey precisely the same meaning for every product in the catalog and be described in a uniform number and measurement format, such as 45.5 CM. Failure to enforce this standard may lead to various interpretations of how to record the data (e.g., 45.5 Cent., 45 ½ Cent, Forty-Five and a half centimeters, etc.). 

This inconsistency is terrible for website filtering and SEO, and also looks sloppy to customers. Maintaining this level of consistency becomes even more challenging when dealing with a larger number of suppliers, each with their own naming conventions and data standards; AI can prove incredibly helpful in bringing order to this potential chaos. 

3. How will you know your product data is accurate? 

Your product data may be complete and consistent, but this doesn’t mean it’s correctly describing your product! Accuracy refers to the precision and correctness of the underlying data. Getting it wrong can lead to an irritated consumer, an expensive return, or a lawsuit in extreme cases. Ensuring accuracy can be difficult, but here are a few options that can help: 

1. Establish rules and validations: If you sell shoes between sizes 4 and 12, any number greater than 12 or less than 4, including negative numbers or fractions, can be ruled out programmatically. 
2. Verify with 3rd parties: If other sources maintain the same data field, cross-reference it to identify any discrepancies (suppliers or even competitors can be useful).
3. Human checks: Ensure one or even multiple employees check product listings as part of an approval process before they go out to customers.
4. Create a feedback loop: Mistakes will be made, but the faster you can find and fix them, the less damage they’ll do. Leverage your customer support teams, user generated content, or even sales data to quickly find product information errors. 
5. Audit: Routinely carry out audits on small sections of your product catalog to understand its health. Many accuracy issues are systematic meaning they can be pretty easily identified and quickly fixed.

4. Do you have enough product data? 

A data science purist may roll their eyes at this one, but I’m including it based on past experiences. When working with brands, one of the first inquiries I make is around the condition of their product data and the types of errors and issues they encounter. 

Every brand has data quality issues, and most readily admit to it when prodded enough. However, I consistently come across brands that insist they have perfect data. In this instance I almost inevitably go on to find  that the reason their data errors are so scarce is because their product information itself is so scarce! 

Think of it like performing a pencil dive during an Olympic diving competition. Sure, you may not technically have committed any errors, but you’re not getting a 10.0! These brands are leaving their customers in a product experience desert, a situation particularly common with B2B distributors and fast fashion retailers.

It’s easy to govern your data when you’re only providing a Product Name, Main Image, Short Description, and a handful of product attributes. But this doesn’t contribute positively to the customer experience. That’s why one of the most critical questions to consider when revamping your product data is whether there’s enough of it. A properly built product listing should include well over a dozen product attributes, with more complex products having 100 or more. 

At the minimum, you should ensure that you’re providing customers with:

1. Technical Specs (width, height, weight, tolerance, materials, voltage, etc.)
2. Marketing Copy (product/brand/designer story) 
3. Assets (Images, video, audio, manuals)
4. Usage (instructions on how to use your product)
5. Regulatory compliance (environmental impact of product, regulatory compliance etc.)
6. Compatibility & Cross/Upsell (what other products are similar or work well with this one) 

5. How will you handle future changes to your product data? 

Your product catalog is never complete. There will be new SKUs, new product types, new ways of describing products, new suppliers, new places to sell, and an endless stream of directives from the powers that be to change the way you operate. It’s critical that each one of the controls you implement as part of your data governance policy have a mechanism to incorporate change. 

Document all governance policies from the beginning and determine who’s responsible for incorporating future changes, the process for making these changes, how frequently you’ll make these changes, and what your succession plan is should employees leave. 

Master Data Governance for Your Products 

Data governance doesn’t have to be a complex, expensive proposition when it comes to your product data, but it also cannot be ignored. It is what ensures the experience you provide your customers is error-free and up to standards. 

If you’re implementing a PIM solution or revamping your product catalog, be honest with yourself and the state of your catalog by reviewing the questions we just covered. You may not be able to fix everything at once, but acknowledging and documenting the problem is the first and most difficult step. Changes can then be implemented incrementally over time as you juggle competing priorities. 

If you’re ever unsure of where to go in your data governance journey, take a step back and consider your customer and what changes will have the most positive impact on their product experience.  And if you’re looking for some support on improving or implementing your data governance policies, feel free to reach out to a PX expert today.