July 11, 2023

Optimizing Sales Data for Retail Decision-Making Success



Jay Tian
Product & Data

Next arrow

It may seem like a no-brainer, but sales data is critical and often fraught with small errors that have a big impact on reporting and therefore decision-making. Here are the areas of sales data to review with a fine-toothed comb.

This is the second in our series on retail data best practices. To see our views on product, click here.

SKU Management

Although an old problem for retailers, getting inventory just right continues to be a major challenge leading to inaccurate inventory, allocation miscalculations and missing-in-action items. Below are common nuances we see and how to best handle them. 

Accounting for Bundled Inventory

With the popularity of personalization and curated experiences, subscription box services have flourished in recent years. In addition to boxes, many retailers are combining items in their storefronts (in-person and virtual) to give customers a head-to-toe look or many variations on a favorite item, like a t-shirt, long-sleeve tee and sweatshirt. These packaging decisions add value for customers and complexity to inventory management. 

For more on bundle verbiage and nuance, see our article here

Our customers often ask how to best plan for bundles: Do you plan at the component level or final bundle? We tend to answer this question by asking, “How do you hold your inventory?” If the data is housed in your systems at the component level, that’s how we’ll need to plan. 

There are various ways this data might be structured, depending on the systems in use and capabilities of those systems to support this use case, but the most important thing for the scope of this article is to acknowledge the nuance and emphasize the importance of collecting use cases for this data to decide just how to structure it.

SKU Variation

SKUs are always unique identifiers, right? Well yes, but often in a long-running, high-performing SKU small adjustments are made. Maybe that best selling shoe uses a specific thread that gets discontinued or has a too-long lead time to continue using. To the consumer, it will look the same and their demand will be the same, to the point where identifying it as a whole new SKU would simply be silly. Yet, the manufacturer will still need to know which thread to use. So, what’s a retailer to do? 

The answer lies in SKU aliasing or Master SKU logic. If you’re using a tool like Toolio, we help enable this (more here), but if you’re not, you can still acknowledge this need. That’s always the first step. And then determine how your team will manage the SKU on an ongoing basis.

SKU Creation

Simply creating a SKU for the first time can be a cumbersome process for some retailers. Again, tooling can help alleviate this pain (like Toolio), but if you don’t have tooling, here are steps to take to ensure SKU creation is consistent.

  1. Refer to the Product Attribute section of this article.
  2. Establish a clear and consistent naming convention for SKUs. This convention should be logical, easy to understand, and scalable. It may involve incorporating elements such as product category, attributes, and unique identifiers. Document the naming convention and provide guidelines for the team to follow.
  3. Create a SKU creation template. Whether automated or manual, a standardized SKU creation request form or template should capture all the necessary information about the new product, including its attributes, pricing, packaging, and any other relevant details. 
  4. Review and approve new SKUs. Implement a review and approval process for new SKU creation requests. Designate individuals or teams responsible for reviewing the requests and ensuring they comply with the established guidelines. This step helps maintain consistency and accuracy in SKU creation.
  5. Manage unique identifiers. Establish a system to manage unique identifiers for SKUs. This could involve assigning a unique SKU number or barcode to each new product. Implement protocols to avoid duplication of SKUs and ensure they are properly linked to corresponding products in inventory management systems. This also relates to the Master SKU concept above.
  6. Document and train. Provide comprehensive training to the operations team on the SKU creation process. Develop documentation, guidelines, and training materials that outline the steps, naming conventions, and best practices for SKU creation. Regularly update and communicate any changes or improvements to the process.

Finally, create a cadence to review existing SKUs and retire them as necessary. 


Ugh. Am I right? Returns are a trillion dollar challenge with an entire discipline to handle. Therefore the scope of this piece is to simply highlight some tips for ensuring the data flows in correctly.

Managing Returns Data

Returns highlight why managing the overlap between people, process and technology is key. In this section, we’re only focused on situations where you own all three. I.e. you have not outsourced management of your stores and you’re not drop-shipping through a wholesale relationship.

  1. Clear return receipt acknowledgement. Whether it’s a clerk at a physical store or a warehouse associate, the first step is to scan the SKU back into inventory. 
  2. Define condition. Again, who does this task might vary, but as some point, the condition must be decided and populated on the item. At this point, whether it goes back to the shelf as available to sell (ATS) or not will determine where it shows up in the data.
  3. Location designation. Let’s assume the item is ATS, that’s where allocation picks back up. Where will the item go next?

Of course inventory considerations are only one half of the equation, the transaction information will also need to be accounted for. Some of this will depend on your systems and data structures, but generally, we see retailers handle returns data by:

  1. Maintain initial transaction. The initial transaction record is rarely updated.
  2. Create a return transaction. Instead of adjusting the initial transaction record, a new return transaction is created with a reference to the original transaction. This will allow for the removal of this demand for future planning if desired.
  3. Maintain return transaction statuses. For accounting, it’s critical to understand if a return has been requested, processed and refunded.

Drop Ship Considerations

If you have a drop ship relationship with a major retailer, like Nordstrom or Neiman Marcus, you’re likely familiar with the PDF they send notifying you of a return. Although less-than-ideal, here’s our best practice for handling the data.

Start your process when you get your hands on the item. Wait until the item officially hits your warehouse and follow the above steps. If you’re using a 3PL, make sure they’re tight on this process with their team and technology.

IMU, MSRP, Discounting and Markdowns 

Managing cost, MSRP and all the adjustments made in between is critical, but, like all the things on this list, filled with nuance that deserves some extra attention.

For more on IMU vs. maintained mark-up (MMU), see here.

What terms are used to describe retail pricing?

Our glossary of retail terms is long for a reason. Different retailers use different terms for the same data point. 

  • The cost to produce an item: unit_cost, item_cost, cost of goods sold (COGs), FIFO or LIFO (accounting terms for which costs to recognize)
  • The price on an item’s tag: MSRP (manufacturer suggested retail price), IMU (initial mark-up), unit_ticket, compared_at (Shopify-specific data field)
  • The price the customer paid: retail price, sale price
  • Taking value off an item for good: markdown
  • Taking value off an item for a limited time for promotional, loyalty or seasonal reasons: discount

How should discount and markdown be reflected in the data?

Although it might be tempting to simplify the data to a single discount field, we recommend capturing discounts and markdowns in two separate fields to allow for deeper analysis. If you do make this change, make sure all fields taking these numbers into consideration are also updated, like discount total, gross profit, and subtotal.

How attributes can help

Attributes can help for organizing the data for better understanding around discounting and markdowns as well. We advise customers to think about this in the context of the product lifecycle. If an attribute is created to track the lifecycle stage of a product, planners can assess demand in relation to that status. To learn more, see this article

Gift Cards

Gift cards are a special animal and a particularly thorny challenge for finance. We sometimes see customers bring gift card purchases, meaning the initial purchase of the gift card, not the use of gift cards as tender, into the regular sales feed. This is not advised. Instead, gift cards should have their own object where additional details in terms of expiration and remaining balance are kept. See how NetSuite captures this information as an example here.


Location data is critical for retailers, especially as they expand operations and physical store locations. We explain more details on that below in our allocation section, but in the context of sales, the most important consideration is simply to ensure transactions document sales location. Even if that location is virtual. Knowing specifically which channel a purchase came through is critical for future planning. 


For international retailers, currency can be tricky. Again, the right system will alleviate headaches here with conversion mechanisms and multi-currency features. Toolio for example can integrate with multiple Shopify stores, bringing in currency in tendered form and translating to a single currency for reporting and future planning. 

Next arrow