Why Unica Detect could be the perfect solution for EBM needs
Unica Detect is a tool which helps implement complicated business logic in order to realize Event Based Marketing (EBM) strategies.
What is Event Based Marketing or EBM?
Event based marketing is a concept in marketing where a business is able to detect a change in consumer behavior instead of pushing products on them; thus bringing about a change in the needs of a consumer. Once the need is identified we would be able to target a potential customer at the right time for them to respond positively.
Why Unica Detect?
Unica Detect has various unique features that enable it to be a powerful implementer of event based marketing strategies.
POWERFUL ENGINE: A powerful logic engine, which can analyze the most atomic transactions and complicated set of business rules utilizing optimum resources in a very short period of time.
STORING STATE OF CUSTOMERS: In order to be able to detect a change in the behavior of a customer it is necessary to store historical transactional data of the customer. Unica Detect utilizes a unique method of updating the “state” (historical transactional data) of a customer so that every time it reads any transactional data related to the customer it updates the customer’s “state” in a 32 byte hexadecimal format which consumes much less space compared to storing all these transactional data in a table. This method of storing the historical data reduces the direct cost in terms of database space and times required for I/O operations to store and fetch such data via the traditional methods.
– SIMPLE COMPONENT. This component is used to read data from the source file or feed file and make that data available for the other components.
– CONTAINER MANIPULATOR. This component is used to perform very basic manipulations and insert data into a “Container”.
– CONTAINER. A container is a storage area for all the data read by detect. The data for each customer is stored as the state for that customer. The container possesses the ability to age out older irrelevant data from the state of the customer , it can roll up transactional data across a host of parameters. Containers are reusable i.e. they can be used time and again across various triggers where the same data or a subset of the data needs to be used.
– MATH COMPONENT. This component is used for performing various kinds of mathematical operations on the data. Using the MATH component, any kind of mathematical function can be performed.
– TREND COMPONENT. This component in conjunction with other components allows us to use complicated statistical functions to be able to analyze the historical data or acquire a state to be able to decipher the meaning of certain changes in behavior (or events) in a customer’s transactions. There are three basic kinds of trend components.
1. ESD (Exceeded Standard Deviation). This type is used to detect customers who exceeded standard deviation across a historical transaction period taken as standard; it can detect positive as well as negative deviations.
2. Spike. This type is used to detect a sudden spike in the transaction either volume or load it can detect positive as well as negative spike.
3. Trend. This type is used to detect particular type of trends in the transactions like a steady rise or fall within a specific span of time, this growth or decline can be measured in terms of percentage or absolute values.
– SELECT COMPONENT. This component is used toselect smaller data sets from containers pertaining to a specific period so that required operations can be performed on them.
– FORWARD LOOKING INACTIVITY. Unica Detect can be tuned or taught to be able to predict the change in the behavior of a customer at a future point in time using the “Forward Looking Inactivity Component”.
– BACKWARD LOOKING INACTIVITY. Conversely Unica Detect can also be used to reference past behavior to be able to predict or catch an event which has just happened or is about to happen presently.
PATTERN. This component can analyze the transactions of a customer and catch some patterns of significant events or change in behavior it can also bet used to detect a pattern of recurring patterns. Eg. Account balance of a customer less that x amount three or more times a month and this pattern reoccurs consecutively for two months in the last three months.