Amplitude Data Integration with Snowflake

The Two Paths to Connect Amplitude to Snowflake

There are two ways to sync data from Amplitude into your data warehouse for analytics.

Method 1: Manually Developing a Custom Data Pipeline Yourself

Write code from scratch or use an open-source framework to build an integration between Amplitude and Snowflake.

Method 2: Automating the ETL Process with a No-Code Solution

Leverage a pre-built connector from a cloud-hosted solution like Portable.

How to Create Value with Amplitude Data

Teams connect Amplitude to their data warehouse to build dashboards and generate value for their business. Let’s dig into the capabilities Amplitude exposes via their API, outline insights you can build with the data, and summarize the most common analytics environments that teams are using to process their Amplitude data.

Extract: What Data Can You Extract from the Amplitude API?

To help clients power downstream analytics, Amplitude offers an application programming interface (API) for clients to extract data on business entities.

You can visit the Amplitude API Documentation to explore the entire catalog of available API resources and the complete schema definition for each.

As you think about the data you will need for analytics, don’t forget that Portable offers no-code integrations to other similar applications.

Regardless of the SaaS solution you use, it’s important to find a solution with robust data available for analytics.

Load: Which Destinations Are Best for Your Amplitude ETL Pipeline?

To turn raw data from Amplitude into dashboards, most companies centralize information into a data warehouse or data lake. For Portable clients, the most common ETL pipelines are:

Once you have a destination to load the data, it’s common to combine Amplitude data with information from other enterprise applications like Jira, Mailchimp, HubSpot, Zendesk, and Klaviyo.

From there, you can build cross-functional dashboards in a visualization tool like Power BI, Tableau, Looker, or Retool.

Develop: Which Dashboards Should You Build with Amplitude Data?

Now that you have identified the data you want to extract, the next step is to plan out the dashboards you can build with the data.

As a process, you want to consume raw data, overlay SQL logic, and build a dashboard to either 1) increase revenue or 2) decrease costs.

Replicating Amplitude data into your cloud data warehouse can unlock a wide array of opportunities to power analytics, automate workflows, and develop products. The use cases are endless.

Now that we have a clear sense of the insights we can create, let’s compare the process of developing a custom Amplitude integration with the benefits of using a no-code ETL solution like Portable.

Method 1: Building a Custom Amplitude ETL Pipeline

To build your own Amplitude integration, there are three steps:

Navigate the Amplitude API documentation
Make your first API request
Turn an API request into a complete data pipeline

Let’s walk through the process in more detail.

How to Interpret Amplitude’s API Documentation

When reading API documentation, there are a handful of key concepts to consider.

Authentication

There are many common authentication mechanisms. OAuth 2.0 (Auth Code and Client Credentials), API Keys, JWT Tokens, Personal Access Tokens, Basic Authentication, etc. For Amplitude, it’s important to identify the authentication mechanism and how best to incorporate the necessary credentials into your API requests.

Resources

It’s important to identify the Amplitude API endpoints you want to use for analytics. Most APIs offer a combination of GET, POST, PUT, and DELETE request methods; however, for analytics, GET requests are typically the most useful. At times, POST requests can be used to extract data as well.

Request Parameters

For each API endpoint you would like to use for analytics, you need to understand the method (GET, POST, PUT, or DELETE) and the URL, but there are other considerations to take into account as well. You should look out for pagination mechanics, query parameters, and parameters that are added to the request path.

How Do You Call the Amplitude API? (Tutorial)

Follow the instructions above to read the Amplitude API documentation
Identify and collect your credentials for authentication
Pick the API resource you want to pull data from
Configure the necessary parameters, method, and URL to make your first request (e.g. with curl or Postman)
Add your credentials and make your first API call

How Do You Maintain a Custom Amplitude to Snowflake ETL Pipeline?

Making a call to the Amplitude API is just the beginning of maintaining a complete custom ETL pipeline.

Here is a getting-started guide to building a production-grade pipeline for Amplitude:

For each API endpoint, define schemas (which fields exist and the type for each)
Process the API response and parse the data (typically parsing JSON or XML)
Handle and replicate nested objects and custom fields
Identify which Amplitude fields are primary keys and which keys are required vs. optional
Version control your changes in a git-based workflow (using GitHub, GitLab, etc.)
Handle code dependencies in your toolchain and the upgrades that come with each
Monitor the health of the upstream API, and —when things go wrong— troubleshoot via the status page, reach out to support, and open tickets
Handle error codes (HTTP error codes like 400s, 500s, etc.)
Manage and respect rate limits imposed by the server

We won’t go into detail on all of the items above, but rate limits are a great example of the complexity found in a production-grade data pipeline.

If you don’t respect rate limits, and if you can’t handle server responses (like 429 errors with a Retry-After header), your pipeline can break, and analytics can become out-of-date.

What Are the Drawbacks of Building the Amplitude ETL Pipeline Yourself?

You can probably tell at this point that there is a lot of work that goes into building and maintaining an ETL pipeline from Amplitude to your data warehouse.

If you want less development work, faster insights, and no ongoing responsibilities, you should consider a cloud-hosted ETL solution.

Let’s walk through the setup process for a no-code ETL solution and its benefits.

Method 2: Using a No-Code Amplitude ETL Solution

No-code ETL solutions are simple. Vendors specialize in building and maintaining data pipelines on your behalf. Instead of starting from scratch for each integration. Companies like Portable create connector templates that can be leveraged by hundreds or thousands of clients.

Step-By-Step Tutorial for Configuring Your Amplitude ETL Pipeline

Off-the-shelf ETL tools offer a no-code setup process. Here are the instructions to connect Amplitude to your cloud data warehouse with Portable.

Create an account
Add a source —search for and select Amplitude
Authenticate with Amplitude using the instructions in the Portable console
Select Snowflake and authenticate
Set up a flow connecting Amplitude to your analytics environment
Run your flow to replicate data from Amplitude to your warehouse
Use the dropdown to set your data flow to run on a cadence

What Are the Benefits of Using Portable for Amplitude ETL?

No-Code Simplicity

Start moving Amplitude data in minutes. Save yourself the headaches of reading API documentation, writing code, and worrying about maintenance. Leave the hassle to us.

Easy to Understand Pricing

With predictable, fixed-cost pricing per data flow, you know exactly how much your Amplitude integration will cost every month.

Fast Development Speeds

Access lightning-fast connector development. Portable can build new integrations on-demand in hours or days.

Hands-On Support

APIs change. Schemas evolve. Amplitude will have maintenance issues and errors. With Portable, we will do everything in our power to make your life easier.

Unlimited Data Volumes

You can move as much data from Amplitude to Snowflake as you want without worrying about usage credits or overages. Instead of analyzing your ETL costs, you should be analyzing your data.