BLOGS

Your guide to data mapping

August 4, 2022
4 minute read

As the old saying goes, you can’t compare apples and oranges. Nor can you compare grandmothers and toads, cows and long johns, or honey and butter. Many cultures and languages have their own spin on why it’s incredibly difficult to compare two different things. When working with data, though, we have a solution for this age old conundrum: data mapping. 

You see, data mapping allows us to adjust data from multiple sources so that they align. Think of it as turning those pesky oranges into apples. 

What is data mapping

Data mapping is essentially the process of matching fields between two different data sets. It is a crucial step when you’re combining data, since different data sources could have different styles, formatting, or names for the same kinds of data. 

Take the Funnel office for instance. In Stockholm, our team uses a 24-hour clock. That means the Swedish teams set afternoon meetings at times like 14:00 or 16:15. Meanwhile, the colleagues in our Boston office would say those meetings are at 2 p.m. and 4:15 p.m. 

Our calendar software needs to apply data mapping to ensure that our systems see those 2 different time formats as the same thing. Simplified image of data mapping

Additionally, our Boston team members write the date differently than those in Stockholm. In the States, you might write 08/01/2022 (month/day/year) to represent the 1st of August. However, those in Sweden write 2022/08/01 (year/month/day) for the same date. 

If you were to look at data from 1 August (or August 1), you would need to apply data mapping to ensure that one of these styles is converted to the other. 

Why is data mapping essential?

In addition to ensuring your team is working at the same time, data mapping is a crucial part of data management and handling processes. In fact, there are 4 processes that stand out: integration, migration, transformation, and warehousing. 

Data integration

Integration is the ongoing moving of data from one place (a data source) to another (a destination). Data mapping is a critical piece of integration since the map tells the data which fields to go into at the destination.

Migration

As the name implies, data migration is the process of moving data from a source to a destination. Unlike integration though, which is ongoing, migration refers to a one-time transfer of data from the data source, to the destination. 

Transformation

In data transformation, you convert data from its source formatting to a format that fits its destination. Like in our time and date examples above, data transformation ensures that all styles are uniform. 

Warehousing

For those needing to pool data in a single large source, you may need to employ a data warehouse. Here, data mapping acts as a sort of seating chart, telling any incoming data where it should go. 

What is the data mapping process?

Data mapping can be broken down into 5 main steps:

 

1.  Determine what needs to be mapped

You can’t hope to map your data if you don’t know what you’re mapping. That’s why your very first step in the process is to identify the values and fields that need to be mapped. When doing this, it’s important to think of what your main goals are. If you’re simply integrating or migrating data, your mapping may be different (and less complex) than if you are transforming it. 

 

2.  Standardize your formats

Going back again to our time and date examples above, you need to make a decision as to what the master formatting will be. For the time, will the final formatting be Swedish or American? What about the format of the date? Now is the time to make those decisions. 

 

3.  Build the transformation rules and logic

Ok, here is the part that can be tricky. You need to create some form of rules or logic that will identify the data and make any required changes. Sometimes this step may require custom coding, and sometimes your data hub or platform can do this heavy lifting in the background. 

 

4.  Run a test


Now that you have all of your rules and logic in place, it’s time to make sure it all works. Apply your data mapping to a small subset of your data to ensure that nothing breaks or transforms when it shouldn’t. If you’ve manually built your rules and logic in the last step, you’ll want the help of an experienced developer for this part. 

 

5.  Apply the data mapping 


If your test went according to plan and everything is working as it should, it’s time to apply your data mapping to the reset of your data. If you're mapping for integration, be sure to keep an eye on the system to ensure that any changes to the data sources don’t affect your logic. 

Selecting a smart data mapping tool

If you're in the market for a data mapping tool that can handle your needs, you’ll want to select one that is best suited for your specific requirements. A good place to start is to make an honest assessment of your own experience and capabilities in handling data and raw code. Then, determine how much flexibility you’ll need. 

From our perspective, there are 3 areas that you should keep in mind when looking for data mapping tools and comparing them. 

1. Schedule and mapping automation


If you are intending to apply your data mapping frequently and regularly, you may want to think about a tool that can schedule and automate these functions for you. For instance, you may need to pull sales or marketing data from 3-rd party platforms for weekly performance reporting. The more you can automate, the less manual data mapping is required.

 

2. Fluency in many languages


Fluency in programming languages, that is. Depending on your tech stack, you may need a tool that can “speak” in XML, JSON, Python, and more. Be sure to know the limitations of the software and yourself before making any final decisions. 

 

3. Code or no code


If you’re an experienced developer or data scientist, you may be happy to roll up. Your sleeves and create custom code that supports your every desire. If you’re a digital marketing, though, you may want something that skips the hardcore programming part. Instead, you’ll want to opt for a code-free solution that allows you to create your map in a graphical, drag-and-drop style, hiding all of the complex code behind the proverbial curtain. 

Speaking of digital marketers…

Data mapping for marketing data

If you follow the Funnel blog (thank you, we see you) you most likely are NOT a super code savvy developer or data scientist. You’re a modern digital marketer trying to make sense of this whole data mapping thing so that you can use it in your day-to-day work. 

Not to worry. 

You probably already use data mapping without even realizing it. If you use cookies and other tracking technologies to target (and retarget) your audience, the underlying technology is actually a form of data mapping. 

Then, when you need to collect your data, you’ll need to employ a form of data mapping before sending your full dataset on to a visualization tool. When companies approach us early in their data growth journey, we often see them trying to map and transform data manually. 

This often involves collecting data in a massive Google Sheet with multiple tabs. Loads of rules are then applied throughout the sheet in order to ensure data conformity. While this approach is okay for those first starting out, it can quickly become unruly and lead to breakages. 

As your data offering becomes more mature and complex, you should really start thinking about moving to a stronger solution — something more akin to a marketing data hub. That way, you can collect, organize, store, and share your marketing data.