Server-side Quickstart

Get event data flowing from your server-side app in less than 15 minutes.

This quickstart guide walks through code examples using the Node.js SDK. It will help you get started with our server-side SDKs.

For simple use cases, you can get data flowing in less than 15 minutes by following these three steps:

  1. Install a RudderStack SDK in your back-end/server-side app.
  2. Instrument API calls to identify users and capture user actions (like page views and button clicks).
  3. Connect destination integrations to stream events to business tools and your warehouse automatically.
Sign up for a free RudderStack Cloud account before you get started.

Step 1: Installing the Node SDK

In most cases, installing one of RudderStack’s server-side SDKs in a dev environment takes a back-end developer less than 10 minutes.

How data engineers can get the Node SDK code

  1. Sign in to your RudderStack account and click + Add source in the default Connections view.
Add source
  1. Select Node.
Add Node source
  1. Copy the installation snippet under the Setup tab (this snippet contains the write keyThe write key (or source write key) is a unique identifier for your source. RudderStack uses this key to send events from a source to the specified destination. and data plane URLThe data plane URL is the location where events are routed and sent to the RudderStack backend for processing. You can find this URL in the home page of your RudderStack dashboard. , which ensure events are sent to the correct source).
Node SDK snippet
  1. Send the snippet to your back-end development team.

How back-end engineers can install the Node SDK

Install the Node SDK in your back-end app using npm by running the following command:

npm install @rudderstack/rudder-sdk-node

Initialize the SDK

Your data engineer should have provided you the Node SDK snippet. Follow these steps to initialize the SDK:

The snippet you recieved should already contain the write keyThe write key (or source write key) is a unique identifier for your source. RudderStack uses this key to send events from a source to the specified destination. and data plane URLThe data plane URL is the location where events are routed and sent to the RudderStack backend for processing. You can find this URL in the home page of your RudderStack dashboard. values, which are needed in order to route the event to the correct source in RudderStack.

For reference, here’s what the Node SDK installation snippet looks like:

const RudderAnalytics = require('@rudderstack/rudder-sdk-node');

const client = new RudderAnalytics(WRITE_KEY, {
  dataPlaneUrl: DATA_PLANE_URL

For more information, see the full Node SDK documentation.

Import the initialization snippet in app.js or any other file in your project, depending on your requirement.
See this sample Node app for more information on initializing the SDK.

Step 2: Instrument API calls

All RudderStack server-side SDKs follow a standard event spec. It helps you plan your event data and supports various API calls for tracking events in your app - these include identify, track, screen, group, alias, and reset calls.

Most users start by implementing track and identify calls.

You can use track calls to record user activities like clicking a button, creating a post, or making a purchase. You can describe track events with propertiesProperties are additional contextual information you can add to a track call to further describe the action a user takes. .

Here is an example of a track call named Item Viewed. The properties are price and item_id:

  userId: "1hKOmRA4GRlm",
  event: "Item Viewed",
  properties: {
    price: 19.95,
    item_id: "1234abcd",

identify calls are typically fired when a user performs an identifying action like creating an account or making a purchase. identify calls also associate a known user with their actions (even if previously anonymous). You can describe users with attributes by adding traitsTraits are attributes that describe a user. They can be added to an identify call in the traits object. Some examples of traits include age, gender, or some specific details - for example, a user’s product plan (free, basic, premium). to the payload.

Here is an example of an identify call. The traits are name, email, plan, and friends.

  userId: "1hKOmRA4GRlm",
  traits: {
    name: "Alex Keener",
    email: "",
    plan: "Free",
    friends: 21,

RudderStack does not store or persist user state in any of the server-side SDKs.

Unlike the client-side SDKs that automatically store state information for individual users, the server-side SDKs don’t persist state. Hence, you must specify either userId or anonymousId every time you make any API calls using a server-side SDK to associate events with individual users.

See the RudderStack Node SDK documentation for more information on the other supported API calls.

Step 3: Verify data flow and connect destination integrations

Once software engineers have installed the SDK and instrumented events, the data team can verify the event flow into RudderStack. Then, they can connect destination integrations to forward events to the tools in their stack automatically.

1. Verify data flow for your Node source

Before connecting destinations, verify that events are flowing into RudderStack by checking the Live Events viewer in your Node source.

To see live events, go to your Node source from the Connections view and click the Live Events button in the top right of your screen.

There will be a delay before you see events in the Live Events view. RudderStack doesn’t store any data; it temporarily opens a gateway between the control plane and data plane to show you the live events. The delays usually last only a few seconds but on the RudderStack Cloud Free plan, they can last up to one minute.
Live Events

After a few seconds, you will see events populating the feed.

If you do not see events in the Live Events view, then there is likely a problem with your instrumentation.

2. Connect destination integrations

RudderStack supports 200+ integrations including data warehouses and data lakes, marketing platforms, CRMs, analytics tools, streaming platforms, and more. You can see the full list of supported destination integrations here.

To add a destination in RudderStack:

  1. Click + Add destination in the default Connections view.
Add destination
  1. From the list, select your preferred destination where you want to route the event data. Here are some popular destinations to help you get started:
AnalyticsAmplitude, Mixpanel, GA4
CRMHubSpot, Salesforce
MarketingBraze, Mailchimp
Object storageAmazon S3, Redis
Streaming platformsApache Kafka, Amazon Kinesis
WarehousesSnowflake, BigQuery, Redshift
Data lakes & lakehousesDatabricks, Google Cloud Storage
  1. Set up the destination by configuring the connection settings. For details, see the destination-specific documentation.

Next steps

This section contains some optional, but helpful steps that leverage RudderStack’s most popular features for transforming events and debugging problems.

Add event transformations

One of RudderStack’s most-loved features is Event Transformations, which you can use to operate on the payloads flowing through RudderStack. You can use it for:

  • Sampling or filtering events.
  • Removing sensitive user PII from your events.
  • Enriching events using static logic, an external API, and more.

To use a transformation:

  1. In the left sidebar, go to Collect > Transformations. Then, click Create Transformation.
Create transformation
  1. Select a transformation template from the list depending on your use case. To create a transformation from scratch, click Custom transformation.
Transformation templates contain prepopulated, ready-to-use transformation logic, which you can apply to your events before sending them to the destination. You can also modify the code as per your needs.
Choose a template
  1. Verify if your transformation works as expected by clicking the Run Test button.
Testing a transformation
  1. Click Save to save your transformation.
  2. Go to the Connections tab of your transformation and connect it to the destination you set up above. See Connect transformation to destination for more information.

When you add a transformation and connect it to a destination, RudderStack does the following:

  1. Tracks events at the source.
  2. Applies the transformation logic to your events.
  3. Converts the events in a format the destination expects - RudderStack does this internally and requires no user intervention.
  4. Sends the transformed events to your destination.


RudderStack gives you complete observability into your events and the ability to debug errors that might occur in case of event failures. It provides the Live Events feature, where you can:

  • Verify your instrumentation.
  • Get a real-time view of the events flowing from your sources to the connected destinations.
  • Identify and debug any errors at the source, destination, or transformation level and narrow down the root cause of the issue.

Source Live Events viewer

This is helpful to verify if RudderStack is receiving the source events at all. The payload you see in this viewer is the raw event payload collected from your website.

Live Events
Source live events details

Transformation Live Events viewer

This viewer will give you a before and after snapshot of the event going into your user transformation and what it looks like afterward. It also notifies you about any dropped events or errors during the transformation, along with the details.

Transformation live events
Transformation live events error message

Destination Live Events viewer

This viewer shows you what the payload looks like when RudderStack sends it to the destination. You can also see the detailed error message if the destination returns an error.

Payload to the destination

See the Live Events documentation for more details.


For any questions on using RudderStack, you can:

Questions? Contact us by email or on Slack