The parts of a track
Create a track
1
Open Tracks
In the dashboard sidebar, under Analyze, click Tracks. If no track is open, the
track builder opens. It can also show your recent tracks.
2
Choose the start
Choose Session to start from every session, Page to start from a page visit, or
Event to start from an event.
3
Choose the goal
Choose one of the detected goals, or search all events.
4
Set the audience and the filters
Choose All users, First time, or Returning. Add filters if you want a narrower
group. Then click Analyze.
page_view to purchase_completed. Add filters after you
have read the full graph.
How GrainQL analyzes a track
1
Find the starters and the converters
A starter is a visitor who did the start event in the date range. A converter is a starter
who also did the goal event in the date range.
2
Measure the time to goal
GrainQL measures the time from the first start event to the first goal event of each
converter. It counts only journeys that took 7 days or less.
3
Score each event
For each event, GrainQL compares the conversion rate of the starters who did the event with
the conversion rate of the starters who did not. The ratio is the lift.
4
Find sequences
GrainQL finds runs of two or three events that raise or lower the conversion rate.
5
Draw the graph
GrainQL draws the transitions between events. A transition belongs to a conversion path or
to a drop-off path.
What a track shows
The toolbar shows the start and the goal. It holds the date range, the number of paths in the graph, and Track settings. The row under the graph shows the main numbers. Click Details to open the full panel.
Boosters and Friction show only events with a confidence of 0.9 or more. A lift of 2.0 means
that the starters who did the event converted twice as often as the starters who did not. A lift
is a correlation. It does not prove that the event causes the conversion.
Controls
Press
Esc to clear the track and return to the track builder.
Page and section events
GrainQL gives each page and each tracked section its own node. A plainpage_view node hides
which page the visitor saw. The normalized node names it.
A path that reads
page_view:/home, button_clicked, page_view:/pricing,
purchase_completed tells you that visitors saw the homepage before the pricing page. You can
use the normalized names as the start or goal event, for example section_view:pricing_table
to checkout_initiated.
GrainQL removes scroll events, heartbeats, and heatmap clicks from the graph. Use heatmaps to
read scroll behavior. Read Heatmaps.
Common tracks
To compare two periods, two variants of an A/B test, or two groups of users, open the same
track with a different date range or different filters.
Read the result
To ask why a number changed, press
Cmd+J (Ctrl+J on Windows and Linux) and ask Neo.
Rules for tracks
- Start with a high-level journey, for example landing page to sign-up.
- Create tracks for the journeys that matter to the business first.
- Do not create a track with narrow filters. It holds too little data to read.
- Do not ignore the drop-off paths. They show where visitors stop.
- If you want journeys across devices, identify your users. Without a user ID, GrainQL cannot connect the events of one person on two devices. Read Identify users.
Limits
A track runs when you open it, not in real time.
What tracks do not do
- A track does not update in real time. It runs when you open it.
- A track does not connect the events of one person across devices without a user ID.
- A track does not show scroll events. Heatmaps do.
- A lift does not prove cause. Test a change before you act on it.