Data migration is often the greatest challenge in the eCommerce replatforming process, particularly for enterprise brands. 

Improperly migrated data sets with complicated merchandising (products, SKUs, categories), mountains of customer data, long order and discount code histories, and more can lead to an inability to run useful (or any) analytics post-migration.

What follows is not a walkthrough about how to migrate eCommerce platforms, as there are some comprehensive resources out there, such as this guide from Shopify and this guide from BigCommerce. 

Rather, this is an overview about migrating your data when you replatform and ensuring that your reporting and analytics remain uninterrupted and unharmed. This article is geared for non-technical readers, in order to explain challenges of the process (as well as a solution for the challenges).

The Two Data Migration Options

Fundamentally, there are only two data migration options while replatforming, though variability within those options exists, depending on your particular data architecture, reporting, and analytics setup.

Also fundamentally, both options require data mapping.

What is Data Mapping?

Data mapping enables you to match a field from one system to another so that you can identify where the same data is stored in two different systems.

For instance: if in your original eCommerce platform, order dates are stored in a table called created_at, and in your new platform, order dates are stored in a table called processed_at, code must be written to ensure that created_at data is transferred and organized under processed_at. 

Shopify, BigCommerce, Magento, Salesforce Commerce Cloud, Solidus, and other eCommerce platforms have different tables, different data models, different APIs, different ways to label data, and different functionality.

Therefore, data between your original platform and new platform will not be 1-1. The data can be organized and labeled in unique ways, and certain functionality on one platform may not be available on another platform. 

This is why when you’re replatforming, a lot of mapping is required. Especially because, in practice, the data mapping is significantly more complicated and numerous than the created_at vs. processed_at example above.

Option 1: Upload All Data from Original Platform into New Platform

When executed perfectly, this option will yield a perfect result: all your original data will be in your new eCommerce platform, and your reporting, analytics, and visualizations will look exactly the same in your BI/analytics tool.

Many brands choose this option because they do not want interruptions to their reporting, and they do not want their customers to know that anything changed. 

However, the process is challenging, especially for brands with a lot of data and/or complicated data, because of (you guessed it:) data mapping. When data is improperly mapped between the platforms, data disasters ensue.

The Outcome You Don’t Want: Data Is Improperly Loaded into the New Platform

Incorrect data mapping will ruin reporting in your new platform. For instance:

A common mistake is when order dates are improperly mapped into the new platform (based on differing fields in our data mapping example above). As a result, the platform may show years’ worth of orders on incorrect dates or even show years’ worth of orders on one single day: the day that the data was migrated. 

That second one is extreme, but it happens, unfortunately: in the new platform, in all your reporting, and in all your analyses, it might show 40 million orders that occurred on one day.

However, even with some incorrectly mapped data, all historical data will still be ruined.

This subsequently ruins any metrics and analyses: no more SKU performance, LTV, cohort analyses, RFM, or repurchase analyses. From a data collection standpoint, you’ll be starting from square one.

Option 2: Keep All Data in Your Data Warehouse, Do Not Upload Data into New Platform

If you have a data warehouse, we recommend building your new reporting and analytics by blending data that currently exists in your data warehouse from your original eCommerce platform, so the data that will start flowing in from the new one.

• If you don’t have a data warehouse, such as Snowflake, we highly recommend getting one as soon as possible, to store and centralize all your data—including your orders data.

In this situation, you’ll maintain your historical order data in your warehouse, switch platforms, and lose orders data in the new platform, but you’ll be able to keep all your reporting and analytics intact because of your warehouse data.

In short, at the very least, your original data will be safe.

However, as in Option 1, the data mapping challenge remains. 

It is difficult to combine the data from the two platforms, and you will need to map all your data that was loaded into your data warehouse from your original platform and unify it with the data from your new platform. 

Option 2.5: Use Daasity to Normalize Pre- and/or Post-Migration Data

Okay, we lied a bit. There’s another option, but it’s an expanded version of option 2.

We’re Daasity, a data and analytics platform that is purpose-built for consumer brands. We work with multi-million and multi-billion dollar brands and enable them to be truly data-driven organizations by building a single source of truth around their data, automating their reporting, and enabling them to run deep analytics. 

If you’re looking to create a seamless reporting and analytics transition as you migrate to a new eCommerce platform, we can help before you replatform:

If you do not currently have a data warehouse:

We can help you set up a data warehouse and load your historical data there. That way, you can get up and running with reporting and analytics while still using your original eCommerce platform.

Then, when you do replatform, your data will stream into Daasity. All your reports, visualizations, metrics, and analyses will be identical, as if you had no replatforming process in the first place.

It will be like turning off one data tap and turning on another: all your data will be unified.

If you do have a data warehouse:

If you have your own data warehouse, we can help you map tables within your current warehouse to match our data model, enabling you to leverage Daasity for reporting and analytics. 

One of our data models, called our Unified Order Schema, enables you to blend omnichannel data (i.e., your eCommerce, Amazon, retail, and wholesale) data in single visualizations, for holistic analytics and reporting. 

This way, you’ll be able to analyze your historical and newly incoming data from your current eCommerce platform. 

When you flip the switch and move to your new eCommerce platform, your analytics and reporting from before will be uninterrupted, enabling a painless data migration process.

If you’d like to see more about how Daasity works, and how we can help in the eCommerce replatforming process, you can contact us to talk to a data migration expert ASAP.