Definition:
In digital analytics and marketing, a metric is a quantitative value that represents an amount, frequency, duration, proportion, cost or outcome. It is used to describe and compare the behaviour of a website, application, campaign or process over a defined period and within a defined scope.
A metric may be a count, such as the number of sessions; an average, such as average engagement time; or a rate that relates two quantities. Its meaning depends on the definition used by the tool as well as the unit, period and calculation scope.
Metrics are analysed with dimensions, which provide descriptive context. Sessions is a metric; country, device or session source are dimensions that can break it down. Not every metric is a KPI, either. It becomes a key indicator when it is selected to assess a priority objective.
Índice de contenidos
Metrics in digital analytics
In digital analytics, metrics summarise the interactions and outcomes recorded by a platform. They commonly appear as columns in a report, while dimensions organise rows or support segments and filters.
A value can only be interpreted when its calculation is known. Two tools may use the same name with different rules, and an implementation can change the scope, attribution, filters or currency. Definitions therefore need to remain attached to the data, and comparisons need equivalent periods.
Google Analytics maintains a reference for dimensions and metrics with the name, category and description of each available field. Consulting it helps prevent a name from being interpreted differently from the definition actually used by the platform.
Metric examples
The appropriate metrics depend on the channel, business model and question being asked. The following are common in website and application measurement:
- Active users: Distinct people who have interacted with the website or application according to the platform’s criteria.
- Sessions: Periods used to group a user’s interactions.
- Views: The number of pages or screens shown, including repeated views.
- Event count: The total number of times an interaction defined as an event is recorded.
- Engaged sessions: Sessions that meet the engagement criteria established by the tool.
- Engagement rate: The proportion of engaged sessions among all sessions.
- Key events or conversions: Actions that the organisation has marked as relevant to its objectives.
- Revenue: The monetary value recorded for purchases, subscriptions or other configured sources.
An outcome may require several metrics to be understood. For example, an increase in sessions does not by itself indicate an improvement if relevant actions, revenue or traffic quality decline.
Standard, custom and calculated metrics
Standard metrics belong to the platform’s data model and have a predefined meaning. They support comparisons within the same tool, although fields with similar names in other systems should not automatically be treated as equivalent.
A custom metric uses a business-specific numerical value that the tool does not provide as a standard metric. In Google Analytics 4, it can be registered from a numerical parameter sent with an event. Its unit, scope and collection method need to be defined first to avoid ambiguous or duplicate values.
Calculated metrics are produced by applying a formula to other metrics. A conversion rate, margin or average revenue per user may be calculated in this way, provided that the numerator and denominator refer to the same period and scope.
Classifying digital marketing metrics
Grouping metrics by the stage or outcome they describe makes a report easier to read without mixing different signals. The same metric may appear in more than one category when it answers different questions.
Acquisition
Acquisition metrics describe the volume and cost of attracting traffic or users. They include sessions, new users, impressions, clicks, click-through rate, cost per click and cost per acquisition. Channel, source, medium and campaign are dimensions used to break down these values, not metrics in themselves.
Engagement
These metrics help explain what happens after arrival. Common examples include:
- Views per user: The relationship between recorded views and users during the period.
- Average engagement time: The time for which content remains active or in the foreground according to the tool’s measurement rules.
- Engagement rate: The proportion of sessions that meet the engagement criteria.
- Event count: The volume of recorded interactions, which can be broken down by event name.
Conversion and profitability
Outcome metrics connect digital activity with actions and monetary value. Their selection depends on the objective: a retailer may examine purchases, revenue, conversion rate and return on ad spend, while a service business may analyse valid enquiries, opportunities and cost per lead.
The conversion rate needs to identify the action counted as a conversion and its denominator. The same care applies to averages, percentages and ratios. Comparing results requires a consistent formula, period, attribution model, filters and currency.