How to migrate from Segment to Data Pipelines: 4 steps for success

To fully harness the potential of your customer data, you need a customer data platform (CDP) that’s both flexible and powerful. That’s where Data Pipelines comes in.

Whether you’re seeking scalable pricing and more control or are simply looking to consolidate your tech stack, Data Pipelines can help you optimize your data-driven marketing efforts. After all, your customer data is only as powerful as the system that manages it.

Ready to make the switch from Segment to Data Pipelines? Let’s go through the step-by-step process for a seamless migration.

The migration plan

First things first, migrating from Segment to Data Pipelines is not a simple copy-and-paste job. We should know—this migration guide is based on our own experience of moving twelve years of data from Segment to Data Pipelines (more on that below). Yet, with careful planning and execution, the outcome can be extremely positive. 

To help you navigate the process of moving systems, we’ve broken down the migration into four key steps:

1. Audit your existing sources and destinations
2. Connect your sources
3. Add your destinations
4. Test and validate your implementation

Let’s look at each step closely and identify the requirements for a successful migration.

Pro tip: Follow a standard release process by working within a developer environment. That way, you can validate and test your setup before moving into production.

Step 1: Audit your existing sources and destinations

Before diving into your full-blown migration, take stock of your existing sources and destinations. Sources are the websites, cloud applications, or servers you capture data from. Destinations are where you send that data, like your marketing automation platform, support tool, or CRM. 

By default, Data Pipelines will send your source data to all destinations connected to the source. So, if there are any connections that your business is not actively using, it’s best to remove them from your migration list. 

By prioritizing only the data essential to your marketing activities, you not only streamline the process but reduce unnecessary costs. That’s helpful for you and your bottom line.

Here’s an example tracker sheet you can use as you conduct your migration audit:

Pro tips: 

• Review Data Pipeline’s catalog of supported sources and destinations as you go through your audit. Not seeing the one(s) you need? Reach out to product@customer.io to request new integrations. 
• Prefer to be more hands off or simply looking for pro support during the migration? Customer.io’s service partners can help with that. Get matched with the right one.

Step 2: Connect your sources

Segment and Data Pipelines source calls are identical, which makes this part of the migration relatively smooth. There are five core types of sources you can configure within Data Pipelines:

• Website sources, such as JavaScript
• Server sources, such as Node.js, Python, or Go
• Database sources, such as MySQL
• Data warehouse sources, such as Snowflake or Amazon Redshift
• Custom API sources, such as your own hand-rolled integration

Here’s the general process for adding any non-custom sources to Data Pipelines:

Websites and servers	Databases and data warehouses (reverse ETL)
Step 1: Connect your source Step 2: Name your source Step 3: Install snippet Step 4: Check the connection	Step 1: Connect your source Step 2: Sync your source Step 3: Query your source

Pro tips:

• If you’re using Segment’s JavaScript snippet, replace it with the Data Pipelines snippet. Remember, with a JavaScript snippet, you cannot simultaneously send data to Segment and Data Pipelines. So, once you replace the snippet, your data will start flowing to Data Pipelines immediately. 
• With server sources, you have two options: swap out Segment’s server package for the Data Pipelines library or modify your code to send events to both Segment and Data Pipelines simultaneously. The latter approach will take longer, but it allows you to review each source individually. This may reduce the chances of human error throughout the migration.

Step 3: Add your destinations

Now that you have connected your sources, you’re ready to set up your destinations. Of course, the steps involved in configuring your destination will change based on the specific destination you enable. But the process will generally look something like this:

Destinations

Step 1: Choose your destination Step 2: (Optional) Select which sources you want to connect to your destination Step 3: Connect your destination Step 4: Review your destination

The most important aspects of each destination you’ll want to note are its configured Destination Actions in Segment. (We call them Actions in Data Pipelines.)

Actions defines how source data is mapped to your destination. Compare the mapping configurations between Segment and Data Pipelines and translate the values from one tool to the next (they should be similar). While this can be a manual process, it’s critical for maintaining data integrity during the migration.

Pro tip: While you can send any properties and events you want, taking advantage of Data Pipelines’ default event mapping options can save you time and effort. Often, the default configuration choices will be all you need.

Step 4: Test and validate your implementation

Thorough testing is critical to a successful Segment to Data Pipelines migration. Let’s face it: human error can, and probably will, happen. So, following an end-to-end testing process will help mitigate mistakes as you go.  

You can use Data Pipelines to test each configured action. If the data flows correctly, you can then move on to the next action on your list. If an error pops up, you can go back and double-check how you’ve configured the event to find any issues. From there, you can retest and revalidate the action before you move to the next one.

Pro tip: Having a backlog of your changes can be helpful if discrepancies arise later. Use your existing audit sheet to note the timing and changes made during each event test. You can easily reference these notes to see what was modified to determine the next steps.

See it in action: How we seamlessly migrated 12 years of data from Segment to Data Pipelines

If it ain’t broke, don’t fix it. Right? When it comes to your martech stack, a set-it-and-forget-it approach might not be the most optimal. It can lead to tech bloat, haphazard internal processes, and missed opportunities for innovation. That’s not great for business or your customers. 

At Customer.io, we’ve spent the last 12 years collecting customer data to inform our marketing strategies. In 2023, when we launched Data Pipelines, our customer data platform (CDP), we decided to reassess how our martech stack was helping us leverage data in our messaging.

We discovered that while our existing martech ecosystem was serving our needs, we could do so much more if we combined our CDP and marketing automation platform (MAP). It was time to migrate from Segment to Data Pipelines.

Of course, switching solutions came with its challenges. But, after careful planning, many meetings, and a lot of collaboration between our product, marketing ops, and data teams, we completed the migration. The results?

• More efficient marketing operations with one solution to manage our MAP and CDP
• Enhanced collaboration between marketing ops, product, and data teams
• Greater focus on data innovation and strategy

One solution to manage them all

Personalization is a critical component of our marketing strategies at Customer.io—it’s a core aspect of our brand ethos. But, like most customer-centric companies, our ability to personalize messaging is tied to the accessibility of our customer data. That’s where a single solution for messaging and data has made all the difference.

Bonus? Because we automatically integrate Journeys with Data Pipelines, the team didn’t have to migrate the original API call setup from Segment for tracking users; this saved a ton of time during the destination mapping phase.

Teamwork makes the dream work

The migration didn’t just enhance our marketing; it helped foster deeper synergies between our marketing ops, data, and product teams. The best example of this took place during the final phase of our migration. 

As our Principal Product Manager, Sam Nagourney, began configuring the Actions for each destination in Data Pipelines, there would be moments when errors would arise, like an automation trigger no longer firing. By keeping open lines of communication with marketing and data stakeholders—and a detailed backlog of each configured change—every issue was addressed swiftly and collaboratively.

Now, it’s about leveling up our data

With the migration complete, the team’s focus has shifted to refining and optimizing our data strategy to drive further innovation. The first step? Explore opportunities to add new sources and destinations, like reverse ETL, for even more insights into our customers. 

Get started with Data Pipelines

Are you ready to unlock the full potential of your customer data? Make the transition from Segment to Data Pipelines and see how having your CDP and MAP together can drive efficiency, collaboration, and added control. 

Ready to get started? Try Data Pipelines now!