Tracking events

A strategic tracking plan is the first step toward gathering clean, actionable insights that allow you to understand your users, personalize experiences, and drive growth.

This guide outlines best practices to help you design a tracking plan that aligns with your business goals and captures the right data.

Define your goals

First, define what business success looks like. Map key user actions to your business goals to ensure every tracked event serves a purpose, whether it is increasing sales, improving engagement, or reducing churn.

For example:

• If your goal is revenue growth, you should track E-commerce events and the Goal completed event.
• If your goal is engagement, you should track clicks on banners, articles read, or videos watched using engagement events.

Ask some questions to define which events could help you collect insights about your visitors.

• Need to understand interests?
  Track categories, product, or feature page interactions.

• Want to learn about purchase behavior?
  Track cart updates, checkout steps, and order placement.

• Wish to discover content preferences?
  Track article views, video plays, and clicks on banners.

By being intentional and defining goals first, you ensure you capture the data that truly matters while avoiding collecting irrelevant data.

Map user interactions

Identify which pages, sections, or components users engage with the most to understand which ones are most relevant to your goals.

For example:

• Pages
  Home page, category pages, checkout page, demo request page.
• Sections
  Heroes, banners, announcement bars, product grids.
• Components
  CTAs, forms, modals, cards.

Mapping users' interactions helps you pinpoint exactly how they engage with your website and ensures you capture key actions across the site.

Think long term

Even if certain insights do not feel essential right now, tracking them ensures you are ready for future opportunities.

For example:

• You may not need to personalize by interest today, but capturing it helps you measure the size of interest-based segments and design strategies for it later.

• You may not need more than the cart's total value today, but tracking which products the users add to the cart opens the door to product-level personalization in the future.

Always track with tomorrow in mind. Today's optional data can unlock tomorrow's opportunities.

Choose the right event type

Always think about what you want to learn and how you want to use this data to segment your audience. A good way to connect events and analysis is to make a clear mapping of events and audience criteria.

Let's think about some valuable cases.

To segment users who used a specific discount coupon, you should track the Cart viewed, Cart modified, Checkout started events, which fuel the Cart variable.

When you track these events, you can create an audience like:

cart's coupon is "FREE-SHIPPING"

Or use an event expression criteria to segment users based on their past behavior:

user has abandoned a cart with currency "USD"

Using another example, if you wish to create an audience based on the visitor's interest, you should track the Interest shown event and use a criterion like:

user has shown interest in "personalization"

For more information on audience segmentation, see the Audience section.

Track conversions

Our experiment dashboards rely mostly on the Goal completed event. So if you want to measure conversion rates and run AB tests, you need to pair important events and ensure you're tracking them.

An event like Checkout started tells you when someone starts the checkout process and allows you to segment users based on that action. But only by pairing it with a Goal completed you can analyse the conversion rate of this action. The same applies to Cart viewed, Product viewed, etc.

On another example, if you want to populate the Orders and Revenue widgets of your dashboards, you should use the Order placed event. However, if you are running an AB test and your main KPI is the order conversion rate, you'll also need to configure a Goal completed event for purchases. This is because only this event feeds the Performance per Goal widget, where you can compare how different variants perform against your defined conversion goal.

The same principle applies outside of e-commerce. Imagine you are running a lead generation or content engagement experiment. You might track a Post viewed event every time someone reads a blog article, or a Link opened event when a user clicks on a relevant link. These events allow you to segment your audience for experiences, but they don't reveal the percentage of users who actually converted within that part of your strategy. To capture this, you should pair them with a Goal completed event, such as blog-post-read or download-link-open. With this, instead of just knowing that 200 users opened a link, you can also see that 150 users opened a link across 600 sessions, giving you a clear conversion rate of 25%.

Keep event names stable

For events that require custom naming, like the Goal completed and the Event occurred events, avoid using terms tied to visual aspects or content that might change with time.

For example, instead of calling an event hero-blue-button-click or hero-transparent-button-click, which depend on a visual aspect that may change, you should use hero-primary-cta-click and hero-secondary-cta-click, which describe the actual action and hierarchy of the element.

The same applies to naming events after the label or link of a button, such as shop-now-click. Since copy is likely to change, it's better to describe the role of the element in the user journey. In this example, you could use add-to-cart-button-click.

Clear, standardized naming makes your events easier to maintain, analyze, and scale.

Combine broad and detailed tracking

Sometimes it is important to know if the user progressed in the journey and what exactly triggered that behavior. For example, did they click on the card section? If so, which specific card?

To capture both the big picture and the details, combine general and specific tracking. For example, you might use cards-click to track interactions with the entire section, and card-1-click, card-2-click, and similar events to track engagement with individual cards.

Enrich events with details

The more detail you send, the more you can learn and personalize. Include every property you have, even if it is optional.

For Order placed, add properties like:

• The discount applied to the order.
• The coupon code used in the purchase.

For Cart modified, consider:

• Which category each product belongs to.
• The variant of the product, such as size, color, and style.

For Post viewed event, include:

• The tags that describe the post's content.
• Which category the post belongs to (e.g., "news", "tutorial").

These details may seem secondary, but they transform raw data into actionable insights. With them, you can answer more specific questions like "Which coupon generated the most orders?" or "Which type of content engages users the most?" and build sharper audience segments for future experiences.

Explore

Check out these links to learn more about our features and concepts: