Journey Discovery Result & Analysis
- Pete Trevitt
Overview
Journey Discovery is primarily an Analytics tool, with some stats reporting. It is powerful and versatile and, as discussed in the Journey Discovery - Introduction and Considerations , can be used in conjunction with a number of approaches to achieve the required analytics. Journey Discovery have many potential analytics components, including Data Transform, Location Map, Date Range, Execution Parameters, Journey Raking.
As such it is likely that analysts will want to be iterative in their approach, perhaps with different Journey Ranks, or using Journey Discovery to add further Journeys to the Location map.
The results of a Journey Discovery output is a list of Micro Journeys that have been found to occur in the data.
A Micro Journey Pattern is simply a consecutive sequence of Locations made by a Customer/Visitor that make a pattern that occurs in data. To be recognised as a Micro Journey, the Locations must be in a single “Interaction” i.e. the gap between any two Locations is less than 30 mins, or in other words, a series of Customer interactions which have less than a 30 minute gap between each one.
IMPORTANT NOTE: When looking at the Discovery Results, these can ONLY be viewed within the scope of that Micro Journey or in other words, you can only look at this grid one row at a time and can in no way assume that these rows are mutually exclusive, they are not. This is due to the concept of sub-patterns that will undoubtedly occur within the data. See section at bottom on Sub Patterns for more info.
Journey Discovery is an Analytics tool that, as we have previously described, has many potential analytics components, including Data Transform, Location Map, Date Range, Execution Parameters, Journey Raking.
As such it is likely that analysts will want to be iterative in their approach, perhaps with different Journey Ranks, or using Journey Discovery to add further Journeys to the Location map.
The following sections will review for functionality available when viewing the results screen with some guidance on when these might best be used.
Discovery Result
After execution of the scoring process, the Discovery Result can be viewed.
It is again important to note that this result is completely dependant on the quality of data that is being captured.
The Discovery Result screen holds a lot of information in a grid view.
In this view:
All columns apart from Id, can be sorted and moved using the drag icons
The gird counts contain sections on Percent, Conversion Revenue & Service Cost
The following section explains each metric or column in this grid
Column Name | Description |
|---|---|
Pattern Id | Each Journey Discovery process is assigned a unique Id, rather than a name. This Id is visible on the Homepage and acts as the name for the particular Journey Discovery result set. |
Journey | The Journey label is created in one for two scenarios:
|
Micro Journey | This displays the actual Micro Journey pattern that has been discovered |
Micro Journey Name | The Micro Journey label is created in one for two scenarios:
|
patternCount | This count reflect the number of occurrences of that Micro Journey within the data |
sequenceCount | This count is the number of Interactions that the Micro Journey appears in |
Customer Count | This count is the number of Customers or Visitors who performed this Micro Journey |
Length | This count of the number of Locations in the Micro Journey |
Score | When Journey Ranking is used, this is the total score of all the Locations in this Micro Journey |
Percent - Primary Journey | This is a percentage of the Locations in the Micro Journey that are part of the highest ranked Journey |
Percent - Secondary Journey | This is a percentage of the Locations in the Micro Journey that are part of the second highest ranked Journey |
Percent - No Journey | This is a percentage of the Locations in the Micro Journey that are not mapped into any Journeys |
Conversion Revenue - Count | This is a count of Conversion events in the Micro Journey |
Conversion Revenue - Total | This is the Total Revenue generated by Conversion events in the Micro Journey |
Conversion Revenue - per Micro Journey | This is the average revenue for each occurrence of the Micro Journey |
Conversion Revenue - per Event | This is the average revenue for each occurrence of a Conversion event in the Micro Journey |
Service Cost - Count | This is a count of the number of Service Cost values assigned to events in the Micro Journey |
Service Cost - Total | This is the Total Service Cost values generated by any events in the Micro Journey |
Service Cost - per Micro Journey | This is the Average Service Cost value generated for each occurrence of the Micro Journey |
Service Cost - per Event | This is the Average Service Cost value generated for each occurrence Event in the Micro Journey |
Filtering and Searching on the Grid
The grid can be filtered by Journey using the multi select filter option at the top of the column. This allows the focus on a specific Journey, and can be useful when used in conjunction with the Journey Scoring to allow different significant Locations to define the naming of the Journey for a specific Locations.
Locations can also be filtered to give a much more investigative option for Journeys.
In this option the view of Micro Journeys can be filtered to only those that contain only the Locations defined in the query. The UI allows a OR combination when two Locations re selected in the same query element using the multi sect option, but can also be sued as a AND combination by using two separate query elements.
This example shows a query to select Micro Journey that contains EITHER Email/Appt Reminder OR Call Center.
From the filter above these two patterns might be seen, as they contain one or the other
This example shows a query to select Micro Journey that contains BOTH Email/Appt Reminder AND Call Center.
From the filter above these two patterns might be seen, as they contain both Locations
This functionality is useful when approaching from an Unguided or Partially Guided starting point. The Analyst can search for Micro Journeys that contain Locations they are interested in, and use this process to define other Locations in that Journey and create them within Journey map ready for a new execution of Journey Discovery.
Link out to Journey Visualisation Sankey
As described in Journey Discovery - Introduction and Considerations , the rows on the Discovery Result screen display Micro Journeys, and with the consideration of sub patterns, these Micro Journeys are displayed as single entities, not merged Customer/Visitor interactions.
These can however be viewed as overlaying behaviours, by linking out to the Journey Visualisation Sankey.
The Journey Visualisation Sankey also allows the display of the Revenue, Service Cost and Satisfaction totals against the Journey, rather than, when viewed in the Journey Discovery, simply against the single Micro Journey. As the sub pattern display has now been converted to overlaying behaviours, the Conversions, Service Cost and Satisfaction are only count once, so represent the values for the Journey Locations contains in the Journeys selected.
This functionality can be launched in a number of ways:
Display all Micro Journeys in a specific Journey
Either select the Journey the from the Journey column (using CTRL+ launches in a new tab) or highlight one Micro Journey from that Journey and use the Journey link in the top right of the grid.
Display a Manually Selected Set of Micro Journeys
Single Micro Journeys can be highlighted or multi highlighted in the grid. As more rows are select the two counters in the top right of the screen display the number of Micro Journey Patterns and Journeys selected.
IN the screenshot below, 5 rows ahve been selected, which is 5 Patterns, but from 3 Journeys
Clicking on either the Patterns option or the Journeys option will display either
Patterns- Launch a Journey Visualisation Sankey displaying the Micro Journeys selected only
Journey - Launch a Journey Visualisation Sankey displaying the all the Micro Journeys from the selected Journeys.
When a Journey Visualisation Sankey is displayed, it is important to remember that this is a Sankey created based on the Journey Discovery. As the Journey Discovery has already been executed, the Event Stream and the Date Range have already been defined as part of the Journey Discovery process, so these options are not available.
All other filtering however, can be carried out on the Sankey to give more insights into behaviours of Journeys, perhaps on Customer attributes to display differing Journey behaviours by Demographic.
IMPORTANT NOTE ON SUB PATTERNS:
When looking at the Discovery Result these can ONLY be viewed within the scope of that Micro Journey or in other words, you can only look at this grid one row at a time and can in no way assume that these rows are mutually exclusive, they are not. This is due to the concept of sub-patterns that will undoubtedly occur within the data.
Consider the following Micro Journeys:
The first is a 4 length Micro Journey, the second a 5 length Micro Journey, but the first 4 Locations are the same. To explain in its simplest form, the patternCount for the first Micro Journey is 26, so this Micro Journey has occurred 26 times. The count for the second Micro Journey is 23, but all 23 of these are also counted in the first Micro Journey, where the 3 difference being those who are counted in the first, but did not go to the checkout-page.
There would obviously then be another 4 Location Micro Journey being:
which are the last 4 Locations in the 5 Location patterns and this Micro Journey will also have its own row on the grid.
In terms of the Revenue or Cost aggregates, again looing at the above scenario for Revenue, (being the checkout-page) the first Micro Journey would have no Revenue, as there are not Conversion events (checkout-page), the second would have revenue because of the checkout-page Location. Note that the other 4 Location pattern would also have Revenue but, the Revenue value will have been counted in both the 4 and 5 Location length patterns, even though it is clearly counting the same revenue.
So to summarise again, due to sub patterns, and double counting, each row in this grid and each Micro Journey must be considered as its own entity and can bear no relation to other rows.