Trends analysis

Last update: 2024-10-28
  • Created for:
  • User

The GraphTrend Trends analysis provides valuable insight around the performance of your product or the behavior of your users over time. The horizontal axis of this report is a time interval, while the vertical axis measures your desired events.

Use cases

Use cases for this analysis include:

  • Evaluate product performance: Trends allow you to assess the overall performance of your product over a given period. By analyzing metrics such as user engagement, adoption, or conversion rates, you can identify if the performance of your product is improving, stagnating, or declining.
  • Feature adoption: Trends allow you to understand how users adopt new features or updates that you release. You can determine which features are popular and which features require improvement. This information allows you to make data-driven decisions around what features to prioritize your development efforts around.
  • User behavior: Trends can provide insight into user behavior over time. By examining specific actions that users take, you can identify patterns where users might drop off. You can combine insights from this analysis with Funnel for even more insight around behavior.
  • A/B testing and experimentation: If you run A/B tests within your product, you can use Trends to gauge which tests are the most successful over time.

Interface

See Interface for an overview of the Guided analysis interface. The following settings are specific to this analysis:

Query rail

The query rail allows you to configure the following components:

  • View: Switch between this analysis and Frequency.
  • Events & Metrics: The events or metrics that you want to measure. Each selection is represented as a chart series and table row. Events and metrics cannot be combined in the query; once you’ve made your first selection, the remaining query selections must be of the same type. You can include up to five selections.
  • Counted as: The counting method that you want to apply to the selected events. Options include Events, Sessions, Users, Percentage of users, Events per session, and Events per user. Counted as options are only applicable for event queries and are removed for metric queries.
  • Segments: The segments that you want to measure. Each selected segment doubles the number of chart series and table rows. You can include up to five segments.
  • Breakdown property: Breaks down the chart series and table rows by the values of the selected property. A single breakdown property is supported. The top 20 values appear in the table, and up to ten values can be viewed in the chart. You can hide or expose a row in the chart by toggling the Show hide icon icon.

Chart settings

The Trends analysis offers the following chart settings, which can be adjusted in the menu above the chart:

  • Chart type: The type of visualization that you want to use. Options include Line, Bar, Stacked bar, and Stacked area.

Overlays

Add additional data to the chart. When more than one series is visible on the chart, overlays appear on hover only.

  • Anomaly detection: Runs anomaly detection on the trended analysis. Outliers appear as dots that you can hover over for more information.
  • Trendline overlay: Adds a trendline to the chart which helps depict a clearer pattern in the data.
    • Linear: Creates a straight regression line. Best for simple linear data that increases or decreases at a steady rate. Equation: y = a + b * x
    • Logarithmic: Creates a curved regression line. Best for data that increases or decreases quickly, then becomes more level. Equation: y = a + b * log(x)
    • Moving average: Creates a smooth trendline based on a set of averages. Also known as a rolling average, a moving average uses a specific number of previous data points (determined by your selection), averages them, and uses the average as the point in the line. Examples include seven day moving average or four week moving average. Available moving average options depend on your selected interval and date range.

Time comparison

You can compare the current time period to a previous time period. If you select an option in this menu, every data point receives a similarly colored dotted-lined counterpart. This counterpart represents that same metric in the selected previous date range. Setting this option doubles the number of items on the chart and rows in the table.

Available time comparison options include the previous period, 13 weeks prior, 52 weeks prior, and a Customized date range. If you select Customized date range, additional options appear to let you select the number and granularity. If you select None, the date comparison is removed.

Date range

The desired date range for your analysis. There are two components to this setting:

  • Interval: The date granularity that you want to view trended data by. Valid options include Hourly, Daily, Weekly, Monthly, and Quarterly. The same date range can have different intervals which affect the number of data points in the chart and the number of columns in the table. For example, viewing an analysis spanning three days with daily granularity would show only three data points, while an analysis spanning three days with hourly granularity would show 72 data points.
  • Date: The starting and ending date. Rolling date range presets and previously saved custom ranges are available for your convenience, or you can use the calendar selector to choose a fixed date range.

On this page