Customer: The end-user of CIPA
User: The user of the Customer’s website or mobile App, also is the subject of User Behavioural Analytics.
UBA: The abbreviation of User Behavioural Analytics.
What Is Behavioural Analytics
Behavioural analytics is a subset of business analytics that focuses on how and why users of eCommerce platforms, online games, & web applications behave. While business analytics has a more broad focus on the who, what, where and when of business intelligence, behavioural analytics narrows that scope, allowing one to take seemingly unrelated data points in order to extrapolate, predict and determine errors and future trends. — Wikipedia
Based on the purpose, there are two types of behavioural analytics:
- Marketing: Sales promotion, improving user retention and engagement, and so on.
- Security: Detect attacks and insight thread by identifying anomalous behaviour.
This article will focus on discussion about the first type of behavioural analytics.
Fundamental Functions of UBA
To provide UBA services, there are four fundamental functions should be implemented:
- Cohort Analysis
Cohort analysis is a subset of behavioral analytics that takes the data from a given dataset and rather than looking at all users as one unit; it breaks them into related groups for analysis.
- Funnel Analysis
Funnels track the steps users take through a pre-determined flow and provide a visual representation of how users convert, get off-track, or drop off the path that you’ve laid out for them.
- Create metrics for statistics, ratio or other measurements.
- Trace all actions a user takes and all events occurred during a whole session such as from login to logout. Session tracing can help on identifying:
- Detailed user behaviour such as which page people are likely to stay longer.
- Unnecessary steps in a transaction flow.
- Anomalous action sequence.
The user should be a real person with all the attributes he input when he signed up or updated his profile. A UBA product should also provide customers with the ability to add related attributes that a user did not provide.
An event represents any action or interaction that happened in a discrete moment in time.
Timestamp, Entity performing an action, Attributes of that action are three major factors of an event.
Good and Bad Attributes
- Total downloads.
- Sign up date.
- Membership type.
- Games played.
- Referred friend.
- IP Address.
- Last name.
Above is a sample list of good and bad attributes. To summarize, good attributes can be used to segment people into meaningful groups.
Cohort studies, sometimes referred to as panel studies or longitudinal studies, focus on the activities of a cohort group. A cohort is basically a group of people who share common characteristics or experiences within a defined time-span.
Cohort analysis based on two major aspects. One is segmentation — breaking users into related groups for analysis. Another is comparison — comparing different segmentations.
Cohort can be considered as the essential function of UBA because other analyses such as Funnel Analysis also use cohort analysis to drill down the dataset.
Performing cohort analysis
There are following main steps to perform cohort analysis:
The “design step” frames the question or hypothesis to investigate
For example, — do free breakfast upgrades measurably improve business traveler retention for a hotel chain? If there is a good correlation, then a check-in rule might be instituted to upgrade certain types of guests when room availability hits certain thresholds.
The “cohort step” selects the customers to study,typically a control group and one or more target groups
For example, “select from the EDW a list of business travelers based in the north-east market with at least one stay per quarter over the preceding four quarters”. From this pool frequent travellers, subset a control group that did not receive the upgrades in the first quarter, and a target group that did.
The “study step” compares the control and target cohorts over some time range and dimensional grouping criteria.
Any metrics or target groupings can be used in the comparisons. Example metrics are number of stays, length of stays, revenue, non-room charges, etc. Example dimensional groupings are weekend versus non-weekend guests, guests that had meal charges versus those that did not, etc.
- Wikipedia, “Behavioural Analytics”.
- Aukeman, Mark. “Cohort Analysis — understanding your customers”. edwblog.com.
- Balogh, Jonathon. “Introduction to Cohort Analysis for Startups”
- Balogh, Jonathon. “How to do Cohort Analysis in Google Analytics”
- Andrew Chen. “How to measure if users love your product using cohorts and revisit rates”
- Interana. “Behavioural Analytics Ebook”. www.interana.com