Data is a foundational concept in commerce. It is arguably the most critical topic related to success. This blog series covers the different types of data that exist, the importance of data quality and ways data is prepared and transferred in e-commerce today — so you can make appropriate plans or changes to your relationship with data to build a greater chance of success. (Read Parts I & II.)
Let’s look at data in a more granular form, including the logistical challenges in addressing data inconsistencies and paths available for you to convert and adjust data. We will focus on product data, as that is one of Rithum’s most prominent bailiwicks. If you’ve ever used our software and attempted to build Business Rules or Lookup Lists, and your success rate is less-than-ideal, then this might provide insight into why you run into challenges and give you some tips on how to address this with Rithum and any other partners you may share data with. We will not, however, go into scenario specific illustrations — you can read our blog series on Business Rules for that depth.
Illustrating the Challenges of Poor/Inconsistent Data
Product data is easier to transform when there is consistency in the underlying/originating data. While this is not always possible, we will lay out challenges and complications with very basic examples here.
Imagine you are selling a variety of sports equipment. You will need to identify the right sub-category where these products should exist on the marketplace, so they are easily found when users navigate through the left-side menus. In this example, baseball equipment needs to be separate from football, soccer, hockey, etc.
If the data you are working with has the sport (baseball) defined in multiple different fields:
- Title Field Only
- In an attribute called “Sport” only
- In the description only
- In some combination of the three
In inconsistent format:
- Base Ball
- Base ball
- base ball
Misspelling exist too:
Using this very rudimentary example, you can see how quickly things can get ugly from a logic perspective. Then you need to account for all the different locations, spellings and formats in a business rule or lookup list. Rithum’s rules are customizable and powerful enough to address all of this variation, but the point we are illustrating is that starting with consistent data — in format and location — is always a better choice, and will more easily lead to success in listing and selling. If you could simply look for the “Sport” field and know that it would always be populated with properly spelled and consistently formatted values, then, logically speaking, you would need to reference a single field. In addition, if a rule is still needed, then logic for the rule will be simple to convert the value “Baseball” to “Equipment/Sports/Baseball” for the category path requirement.
Options to Address Data Inconsistency
Path #1: Utilize Business Rules & Lookup Lists in Rithum
You are welcome to build out rules and use lookup lists within Rithum to address these kinds of data inconsistencies. We can help you build some to meet your conversion needs, and we encourage their use. Business rules have a number of benefits:
- Data is converted inbound to Rithum and/or outbound to the various channels we support to meet the channel requirements.
- Time spent will often be focused on channel-specific data requirements.
- Business rule functions are very similar to Excel formulas and easy to understand in that context.
- Data does not typically need to be normalized prior to providing it to Rithum.
The real challenges come into play with highly complex business requirements, the complicated business rules to meet those requirements, and the impact after those rules are written:
- Building rules to meet very complicated requirements can be time-consuming and challenging for someone who is not well versed in logic writing — such as Excel functions, coding, etc.
- Troubleshooting the output of a complicated rule can be difficult.
- Rithum can typically decode what is happening, but it may take longer to read, decode, and identify the problem with the logic written.
- Complicated rules process less efficiently than simple rules and basic mappings.
- Chances are higher you will see unexpected differences in how a product is presented between your own website vs. other marketplaces, simply based on how a rule is written.
- When you experience staff turnover in those who built and maintain the business rules and lookup lists, the next person may take longer to understand them and be prepared to make changes.
Path #2: Modify Data Prior to Importing into Rithum
If you have the resources to modify/normalize the data in your system, it is always advisable. This may not always be an inexpensive or realistic option for all sellers, but if some of the following scenarios apply to your company, then this path may be viable:
- Your brand has a limited number of products.
- You are a seller with limited inventory that you control and which does not change with great frequency.
- You have staff who are capable of normalizing the data with some direction.
- You have software designed to normalize the data automatically.
The benefits you will see are not exclusive to your Rithum integration, though, which makes the proposition of executing this path much more attractive:
- Consistent data will benefit all of your partners, including your own website.
- Simpler rules and less time will be spent in each platform to convert the data for channel or other platform requirements.
- It will be faster to locate data in products in each of the platforms if formatted consistently.
- It will be easier to pinpoint and identify issues with data and logic because rules and lookup lists will be rudimentary.
Path #3: A Combination of Paths 1 & 2
If we are honest (and we are), the most realistic scenario to address data inconsistencies is going to be a combination of data preparation prior to providing it to Rithum and the use of business rules and lookup lists for the more basic business logic. This way, you’ll benefit from the features of Rithum and still have access to benefits gleaned from the data preparation.
In the final post of this series, we will review data delivery and methodology to provide more context on the topic of data communication. After all, understanding the data, how it is cleaned up and normalized is of little benefit if you can’t share it.