Event-driven marketing (EDM) is a discipline within marketing, where commercial and communication activities are based upon the measurement of relevant and identifiable changes in a customer's individual needs. It is a distinct and different approach to direct marketing and traditional database marketing which typically use statistical methods.
In this context, an "Event" is defined as a detectable change in an individual's circumstances, today, which is relevant and significant, either in fact or in their mind.
A "Significant Event" is a major happening in a customer's life. They can lead to a measurable change in a customer's normal behaviour, state of mind, personal circumstance, or interaction pattern. It offers a reason to communicate with the customer, with a relevant proposal, at the right time. Events increase the knowledge, understanding and information about a customer that enables the marketer to make better, more informed decisions.
Event-driven marketing is synonymous with event-based marketing (EBM). In some cases it is also referred to as Trigger-based marketing (although "Triggers" are very much a subset of Events - see types of Event below). The definition above is consistent for all of these terms.
Maps, Directions, and Place Reviews
Overview
The theory behind event-driven marketing is that the most productive message is one that is relevant to the customer and based on what's going on in their life at that moment in time.
EDM expands the relationship with customers by monitoring them (and their personal or commercial situation) on a constant basis and responding immediately to relevant changes in their circumstances. This includes monitoring both their transactional behaviour (as is done in direct marketing) as well as their life or business circumstances beyond the immediate transactional relationship.
The results from implementing EDM are notably spectacular and have a positive effect on several common marketing KPIs and metrics. It is common to achieve average positive customer response rates to communications of 34%. This compares to normal targeted direct marketing response rates of 1-4%.
Other metrics affected include increased Customer Satisfaction and reduced Churn/Retention. An overview of the different results achieved can be found on eventricity's site.
Finally Events (and EDM) play a key part in Marketing, particularly as the key catalyst in what are called "Moments of Truth" (see Moment of Truth (marketing)).
Event Triggered Marketing Video
History
Event-driven marketing was first implemented in Australia in 1995-1996 at National Australia Bank. This project was the brainchild of Ray O'Brian (Teradata) and Fernando Riccardo (NAB).
Their idea was based on the fact that the then current marketing approaches were using limited amounts of data of variable quality and that these returned pitiful results (1-4%). They postulated that, if only they could monitor a customer's activities in a more timely and dynamic way, they could accurately determine any changes and hence customer needs.
Two things enabled this approach to come to fruition, Teradata's, at that time, unrivalled ability to load and process large volumes of data, and NAB providing the source of customer transactions.
The results were so spectacular that NAB decided to maintain secrecy over this new approach, fearing replication and competition from their rivals. However in 1999-2000 the bank had a change of policy and decided to publicise their results. Since that time they have been a major presenter of EDM at marketing conferences around the world.
By 2000 there were still only a very few exponents of EDM. This was due to the fact that large amounts of transactional data were required, which in turn demanded high end and specialised database machines. The other factor was that there were no EDM 'products' in the market and thus each project tended to be a consultative, bespoke implementation lasting years and costing millions. During this period all implementations tended to be in very large multi-national banks.
By 2003 database and server technology had advanced to the level that EDM could be performed on more modest machines and this opened the market up to other implementations. At around the same time, three EDM products were launched:
- eventricity's Timeframe
- SynapseEBM (now part of Conclusive Marketing)
- Unica's Detect (now called IBM Opportunity Detection).
At the same time, Teradata continued with its approach of implementing EDM on a consultative basis. Between 2003 and now, Teradata Australia has achieved notable successes in projects in Australia and the Pacific region.
Companies using event-driven marketing
The following list is not complete but shows those companies who have implemented EDM and spoken about their results at public conferences.
- ABN Amro (Netherlands)
- Banca Antonveneta (Italy)
- Banco Espírito Santo (BES)
- Bank of America
- Commonwealth Bank of Australia
- HSBC
- K&H Bank (Hungary)
- National Australia Bank
- OCBC (Singapore)
- Raiffeisen Bank (Austria)
- Scotiabank (Canada)
- Union Bank of Norway
- Westpac
Providers / solutions
There are two categories of providers of EDM solutions, those that have an actual, deliverable, demonstrable EDM products and those that provide this through bespoke, consultative development.
More recently, several CRM vendors have developed and introduced Real-Time Decision/Interaction software modules to the market which they state deliver Real-Time Events. In reality, they tend to be Trigger-based (which is very much a subset of proper EDM - see the section on Triggers below). There is an article here which discusses this topic in more detail. It is to be noted that although these vendors imply delivering real-time Event-Driven Marketing, none of these Vendors actually mention EDM or Events in any of their marketing information about their products.
The main vendors in this space, as considered by Forrester in their Real-Time Interaction Management (Q3 2015), are listed below:
Lastly listed below are some vendors who claim to be (at least partially) in this space:
Research / results
Originally there was no empirical research into EDM and all information has been derived from conference presentations. In 2009 a large CRM survey was conducted on behalf of EFMA and Atos Origin. The objective was to provide a clear, unbiased 'state of the union' on differing CRM approaches and techniques. It was conducted with 65 Banks from 29 different countries. Below is an excerpt of the research, covering the results obtained from the differing Database marketing techniques used to create targeted Customer leads / contact lists. These included
- Ad-Hoc (lists generated spontaneously without in-depth analysis)
- Data Mining
- Segments
- Models
- Events
- Inbound (selling to the Customer as part of an inbound call).
Questions were asked about the results obtained from Leads generated by these techniques, specifically the metrics for Positive / Negative responses and for Sales / Opt out.
1. What is the percentage of Positive responses you receive for contacts?
2. What is the percentage of Sales you receive for contacts?
3. What is the percentage of Negative responses you receive for contacts?
4. What is the percentage of Opt-outs you receive for contacts?
Event information
The following sections provide more detail about Events and their characteristics.
Event attributes
An Event was defined as a detectable change in an Individual's circumstances, today, which is significant, either in fact or in their mind. The three key words in this definition are: individual, today and significant.
Individual:
Events are assessed against individuals not groups of people or segments. Analysing a customer's data individually, determines their individual levels and values which is used to determine significance (see below). This analysis shows historical levels, trends and patterns that are specific to each particular person.
Significance:
Significance is a test applied to an 'Event' to determine if the fact that it has happened to a customer is relevant and noteworthy. Significance in this context is therefore a measure of the variance from the norm for that particular individual.
Today:
Once a significant Event has been detected the timeliness of customer communication is crucial. In fact, if the company is not able to communicate with a customer within 48 hours of detecting an Event, they probably should not bother. However, the speed of response required varies depending upon the Event itself. Some require a response within seconds whereas others can still be effective for up to 30 days. For Significant Events research has shown that the customer response rate decreases by about 66% for each 24-hour delay. An average response rate of 70% for contact within 24 hours is common. This drops to around 25% within 48 hours and less than 10% in 72 hours.
With the arrival of web applications and real-time technologies, it is now possible to monitor and detect a Customer's activity in milliseconds instead of hours or days. Thus the Today attribute (above) may be more properly rephrased 'Now'.
The criterion of 'Significance' becomes even more important as such a system may often be processing upwards of 200,000 interactions per second - and only a tiny fraction of these interactions is significant (and therefore an Event).
Perhaps, we now should consider categorising Events as Micro-Events and Macro-Events, with micro being such things as "Customer hovered over Button X" and macro being things like "Customer is getting married".
Types of event
Events can be classified into several different types. These include:
- Micro-Events
- Triggers:
- Simple (predicted) Events
- Significant Events
- Behavioural Events and Lifecycle Events
Micro-events
When referring to digital products (websites, APIs, applications), the information that describes how the user interacts with the website or app is recorded as an 'event'. In this definition an event is a single instance of measured behaviour. The event could describe a user's explicit action such as 'Clicked on Playlist' or an implicit action such as 'Moused Over Playlist'.
Triggers
A Trigger is a circumstance that has happened to a Customer today but which is not necessarily significant. It is this lack of significance that differentiates a Trigger from a Significant Event.
Triggers are not very accurate in terms of determining Customer need or responsiveness and as such their use is not generally advocated. However they can be extremely useful when considered in combination with other Triggers or Events (see behavioural Events below).
Triggers can be very valuable for activation of straightforward business processes. Their best use is for actions that are completely automated and which require no human intervention. Thus they can be programmed to fire whenever a defined circumstance occurs, and provide a clear result. In such situations they are a very cheap and effective tool. Examples can include things such as
- Triggering the issue of a new cheque book when the penultimate one is issued
- Sending an SMS offering a new mobile top-up when the current one has 10 minutes left
- Sending an SMS offering an overdraft facility when an ATM max payment is made, etc.
The key to using effective Triggers is that the resulting action is binary, i.e. Situation A results in Action B. They can be very effective for Service based offers.
Thus a Trigger satisfies the standard Event criteria of Individual and Today but fails in terms of Significance. Numerous studies have shown that the trigger based emails can double the response rate of direct marketing communications.
It should be noted that the timescale for communication of a Trigger can be anything from instantaneous to the maximum of a day or so. After this, it is typically not worthwhile.
Simple (predicted) events
Most Events are reactive, i.e. Based on what has previously happened. But with some Events we can predict situations that will occur in the future. For example, a customer will finish his loan in 30 days.
Undoubtedly, this is important to the customer and a good reason for communication. It is hardly a surprise and there is no urgent need to communicate today. In fact we can schedule a call anytime over a two-week period and get the same result.
Thus the event satisfies the standard Event criteria of Individual and Significant but fails on Today. We call this type of Event a Predicted or Scheduled Event. Examples include
- End of contract,
- End of loan
- Significant Birthday and
- any kind of Customer activity for which there is a known, scheduled date.
The timescale for communication of a scheduled Event can be a window anything from 5 to 30 days or even more.
Significant events
A single, significant event tells us something about the customer. It is an excellent indicator of real / potential change. Customers are very responsive and accepting of communication at this time. Examples include
- Large Deposit,
- Salary start / stop, etc.
The timescale for communication of a Significant Event is usually quite short - usually only 1 to 2 days. After this, the effectiveness drops rapidly (typically response rates drop at 60% per day).
Behavioural (super) events
A single, significant event may tell us something about a customer, but sometimes events happen in groups and the sequence and combination of these events can tell us a lot about the behaviour and circumstances of the customer.
For example, a customer may have a large deposit and this would be of interest, but a large deposit followed by a large withdrawal and then a salary stop (within a short period of time) may signify something much more important such as 'Redundancy'.
A super (or behavioural) event is a situation where a series of circumstances happen to a customer within a specific period of time and possibly in a certain sequence.
These circumstances can be combinations of such things as Events, Scheduled Events, Triggers (see above) and even other Super Events. Examples of circumstances that Super Events can potentially detect include
- Redundancy,
- Churn,
- Moving house / job
- Marriage, etc.
The timescale for communication of a behavioural event can be from 5-30 days or even more.
Lifecycle events
Lifecycle events are very similar to behavioural events and can sometimes be detected in the same way. Their value is undoubted, with one Dutch bank claiming that 55% of ALL product sales they made were from customer Events based on Lifecycle changes. Examples include:
- First job
- First apartment
- Getting married
- First child, etc.
The timescale for communication of a lifecycle event can be from 5-30 days or even more.
Source of the article : Wikipedia
EmoticonEmoticon