Dip into the data pool: Any database anywhere
Suppose you are a large retailer and your IT is POS oriented with the transaction database is sitting on an IBM platform. CRM is sitting in Seibel. You want to attribute the next POS sale to a specific offer and send a personalized follow up. In order to do that you need to requisition a change request to get the two pieces of data together (e.g. The sales transaction and the promo code definitions). You can either: do this once and keep it up to date; set up a daily report that you check visually; have someone summarize for you in some Excel spreadsheet for management. I propose you update the database once and then track your contacts from that point on.
Data stream: new transactions. Arrival rate, variable according to the process
On a periodic basis you receive new information in the form of data updates, batch processing, and live data feeds. This new data is the data stream. You can review each update as an individual customer event. When an event occurs, you want to apply what you know about your customer (e.g. the customer’s state, the customer segment, your current inventory levels, your call center load, and your current available promotions) and take the appropriate action. Ideally you have strategies and tactics to orchestrate how you want to react to each new event as they occur.
Any functionality: Cloud computing
Thanks to Cloud computing IT infrastructure is no longer a constraining factor in implementing new software. Cloud computing offers an inexpensive way to distribute and obtain services. You want email services, surveys, CRM, database, Social CRM, SQL query, etc.; whatever you want to do is already or will soon be available on the cloud. Information exchange is as easy as XML tickets and APIs.
This means that you now have the opportunity to tie information together between any systems that can exchange information electronically. It is now possible to monitor for all events. A Publisher/Subscriber model allows applications to collaborate without the need to build synchronization bridges.
The key, of course, is how seamless you can make the designer interface. The designer is the means for the marketer to specify how they want to link event and actions together regardless of what applications or systems are actually involved.
Just-in-time basis: In near real-time, as you need it
Whenever the next batch of data arrives, for every detectable customer event you have an opportunity to initiate or continue a dialogue with the customer. Associating a set of rules for each of these events according to the customer’s current state give you control over how and when to respond to any event. Your response could be to send a Welcome email using CakeMail or dispatch a call to Outbound Sales or Customer Service, or update a database such as Salesforce or Siebel, or add records to a data warehouse. You may want to treat a complaint from a VIP customer differently than an infrequent customer. Whatever you want to do should always be within the context of the customer’s current state and the state that you want to move them to.
The future of marketing automation
This is the future of marketing automation: an orchestra conductor who receives a steady stream of notes that he dispatches to the appropriate musician (e.g. flute, violin, percussion) according the score that he wrote. Marketers will receive events from any system anywhere and will execute the appropriate action using any system anywhere on a just-in-time basis.
