Skip to main content

Getting Started

Welcome to your Web Console dashboard! This is where you will create, view and monitor your experiments.

You're almost ready to start creating your first experiment! But first we have to tell the dashboard how, where and why you want to run your tests.

Creating A Unit

Units are the unique identifiers that are going to be used to generate a variant. For experiments running across different platforms (iOS, Android, web, email, etc.) the unit should be known across all of them. Most likely the authenticated user's user_id.

Have a look at the Tracking Unit section of the Creating an Experiment docs for more information.

To define your first Unit go to Settings > Units and click on Create Unit. All that needs set here is the name of your unit. We recommend that you give it the same name as it's referenced in your code.


Feel free to add a description, so the rest of your team knows what the unit will reference to and why it exists.

Setting Your Applications

When creating an experiment, you will be asked which platforms the experiment will run on.

To set these up, navigate to Settings > Applications and click Create Application.

Similarly to Units, you can name these whatever you want and add a description to define it more for your team. Some examples of application names might be ios, android or www.

Once your applications are defined, you're ready to move on to...

Setting Your Goals

Goals are the names given to processes that you will later track in your code.

To setup your first goal, head to Settings > Goals and click on Create Goal.

The goal name will be used in your code, so we recommend using a keyword-like or kebab-case name. For example, newsletter_subscription, cpu_load_time or bookings.


Tags are useful for searching, filtering and classifying your goals. We recommend prefixing each tag with the tag's type. For example, location:Header, stack:Backend or psychological:Trust.

Creating A Metric

Lastly, you will need to create some Metrics! Metrics are the parameters that will be used to track your goals. They can give you insights about your business, your users' behavior, the performance of your system and more!

To create a metric, head to Settings > Metrics and click on Create Metric.

Each metric needs a name. Some examples of metrics trackable metrics might be Bookers, Bookings or Time to Book. You can also give your metric a description for clarity across your team.

Metric Impact

Next, you need to choose your metric's impact. For this, you have three choices - positive, negative or unknown.

A metric with a positive impact means that if it is increased, it's good for your business. For example, Bookings, Revenue and Subscribers would all be positive metrics.Metrics with a negative impact, on the other hand, are bad for business if they increase in frequency. For example, Unsubscriptions or Account Deletions.Metrics with an unknown impact are used when the effect of the metric could be either positive or negative. For example, Time on Page could be positive or negative depending on whether the user is engaged, versus whether they are lost.

Metric Type

Next up is the metric's type. There are 4 types to choose from:

Goal Unique Count

Goal Unique Count is used when you want to automatically track the amount of unique users (units) who performed a goal. Bookers is a good example of this as you want to track the amount of people who booked, not the total amount of bookings. For that you would choose...

Goal Count

Goal Count is used for when you want to track every instance of a goal being achieved. Bookings is a good example for this one, as a single user may have made multiple bookings.

Goal Time to Achievement

Goal Time to Achievement is the amount of time between the user being exposed to your experiment and achieving a goal. This is tracked automatically by all of our SDKs.

Goal Property

Goal Property is used when you want to use your own properties for tracking a metric. For example, you may wish to track the Revenue made during your experiment. This will be done by passing the revenue as a property when tracking your goal. For example, in the Javascript SDK it would look like this:

context.track("goal_name", { revenue: 99.99 });

Selecting Goal Property in the Metric Type field adds some extra fields to the form. You can learn more about these in the goal property fields section.

Goal Retention

Goal Retention metrics are used to track data when a user has achieved a goal after a set number of days from when they were exposed to the experiment.

For example, you could track whether users who left items in their checkout cart came back to make a purchase after one week (maybe after receiving an automatic reminder email). This could be done by creating a Retention Metric that tracks a checkout goal and setting the Retention Period to 7 days.

Goal Ratio

A Goal Ratio metric tracks two goals and divides one by the other. When selecting Goal Ratio from the Metric type dropdown, you will be presented with two goal selection sections - a numerator and a denominator.

This metric type is useful for tracking data like click-through-rate. For example, you could choose clicked_signup_button as the numerator goal, and visited_landing_page as the denominator goal. This metric will now track the CTR of the signup button on the landing page.

Internally, the variance of these metrics is automatically adjusted using the delta method.

Metric Goal

The Metric Goal is the goal which your metric relates to. This field will provide a dropdown of any goals that you set up in the Setting Your Goals section.

Goal Property Fields

If your metric is of Goal Property type, you will have to fill in some extra information on where the property value will come from and how you want it to be formatted on the dashboard.

Upon selection, you will be presented with 4 more fields.

Source Property

This field defines where the property value comes from in your track call. For example, if you were to call the track method in your code with the following parameters:

context.track("revenue", {
payment: {
provider: "paypal",
revenue: 99.99,
currency: "USD",

And you want to use the revenue made as your metric's source property, you would pass payment/revenue into the Source Property field.

Format String

The format string defines how you would like your values to be presented in the Web Console. Both the values and the mean of your data can be customised this way. The curly braces {} represent the value and you can put whatever you like around it. For example, if you wanted to track your incoming revenue in dollars, you might want your Format String to be ${}. This would cause a value of 1000 to be presented on the dashboard as $1000.

Below the Formatting fields, you can see a preview of how your value will be formatted.

Scale of Value

The scale of value is a multiplier for your value. This is useful if, for example, your revenue is tracked in cents. You could set the metric's scale of value to 0.01 and a revenue value of 100000 cents would show itself as $1000.


The precision option allows you to set a number of decimal places that your value can have. Set this to 2 and a revenue value of 1000.455 will be presented as $1000.46.


Outlier limits force extreme metric values to be within a specified range to reduce variance. There are four options to choose from:

UnlimitedNo outlier treatment is performed and outliers are included in the data as normal.
QuantileLimits outliers to quantiles, where you can choose your own lower and upper quantile limits.
Standard DeviationCap each value to a multiple of the standard deviation of all seen values.
FixedAllows you to limit outlier values to specific upper and lower values.

Property Filters

Property Filters allow you to filter your metrics to make them more specific. Let's take our example track call from earlier:

context.track("revenue", {
payment: {
provider: "paypal",
revenue: 99.99,
currency: "USD",

If you are tracking your revenue, you may want to limit the metric to amounts that were paid in dollars. This can be done by adding a property filter where payment/currency - Is - USD. Now, only amounts paid in dollars will be tracked by this metric.


In this use-case, you might want to rename the metric to Revenue in Dollars.

The Next Step

Now you have your units, applications, goals and metrics set up, you're ready to move on to the next chapter - creating an experiment!