More than ever marketers are under pressure to deliver measurable revenue gains and returns on investment. At the same time they are facing increasingly competitive markets, greater ad-fatigue, higher customer expectations, a proliferation of new media and sales channels, and shrinking marketing budgets - in parallel to the business demand to deliver more with less. As products and services become more commoditised, marketers need to concentrate their efforts on engaging customers and prospective customers with relevant and useful information, taking a long-term view of customer retention and business value.
What is real-time decisioning?
To achieve these aims, marketers must employ advanced analytics to understand how best to optimise their marketing and branding efforts. To this end, many organisations are now turning to real-time decisioning (RTD) technologies to create relevant and bespoke interactions between their brands and their customers. These real-time engines embed analytics and automated marketing decisioning directly within the transaction streams of websites, inbound and outbound call centres and other marketing platforms.
RTD is a hot topic, with Forrester Research predicting it will become the fastest-growing segment of all enterprise marketing segments, growing at 26% annually through to 2013. However, there are a variety of types of decisioning engines, each with it’s own set of opportunities and challenges and as with any emerging technology, there remains much uncertainty as to how best to implement and integrate RTD into existing marketing efforts.
Real-time decisioning in practice
As a general rule, real-time decisioning technologies fall into three categories:
1) Rule-based
2) Product-based
3) Customer-based
But how does a marketer choose which one is most suited to their organisation’s situation? Each of these three types of engine has its unique benefits, as well as limitations, and there is significant confusion as to which methodology represents the most effective solution when applied to the real-world scenarios faced by different businesses.
Rule-based RTD
A rule-based decisioning engine works on a simple ‘if / when’ principle. It is the easiest and cheapest of the three technologies to deploy and is well suited to organisations that seek to automate a well-defined best practice. A rule-based engine deploys pre-defined business rules in order to apply a best practice to a specific isolated event, such as when an organisation suddenly reduces or cancels a regular order.
Rule-based solutions are well suited to organisations with basic decisioning requirements as there are a number of implementation limitations that should be kept in mind. For example, when administering such systems marketers will often need to create and experiment with various rule sets in order to determine which best match their specific requirements. In addition, rule-based solutions generally do not take into account what may have happened immediately before the event, or longer-term issues such as customer satisfaction. Without the ability to take into account such complexities, rule-based engines can potentially do harm to the long-term client relationship by suggesting inappropriate offers.
As such, marketers must ask themselves, “Can I apply simplified best practices without losing sight of longer-term considerations?” Rule-based solutions are therefore best for situations when the number of potential outcomes is very limited, when historical and contextual information is not available or of little use and when tight guidelines can be created to limit collateral damage.
Product-based RTD
With these limitations in mind, the second option for those considering a decisioning engine is a product-based solution. This type of solution is well suited to drive promotional sales within a discrete channel - when there are a large number of potential outcomes or when complex business rules exist. An everyday example of this would be an online retailer offering a ‘customers who bought this also bought…’ function on its site. With this type of technology the analytic modeling process itself is fully embedded within production systems. Business rules are used to limit and guide outcomes to better meet revenue goals. Algorithms are developed and redefined by automated, self-learning applications in real-time. With their ability to sift through a large amount of transactional data and product offerings, product-based decisioning engines can therefore significantly out perform rule-based engines, especially when contextual information is abundant and ever changing.
However, product-based solutions are not well suited to every business as they focus on ‘shallow’ transactional data and tend to perform far less well within broader customer-focused, cross-channel situations. Because they are automated, product-based decisioning engines cannot easily detect and correct data quality errors that may render subsequent marketing activity as meaningless or inappropriate.
Customer-based RTD
With customers increasingly expecting an integrated, cross-channel experience that is highly relevant to them and their organisation, the third approach to real-time decisioning technology leverages individualised marketing strategies to drive customer satisfaction, retention, cross sell and long-term value. As such, organisations must move beyond simply determining which product or service to offer during a specific interaction. More strategically, they need to determine how, when and importantly if each client contact wants to be contacted via coordinated communications across both online and offline channels.
Customer-based engines leverage sophisticated mathematical algorithms that are built offline directly into and across operational systems, making for a more reliable deployment as marketers can test and validate their models in staging areas prior to it going live. Customer-based systems pull data from all systems across all points of contact and are able to take full advantage of information gleaned from real-time events. Unlike the other two types of engine, they also leverage an organisation’s full understanding of each customer, driving long-term customer satisfaction, retention and lifetime value. These systems would therefore be particularly relevant for financial services, telco and retail sectors where a dynamic understanding of the customer and their company transactions would enhance the customer experience and increase levels of satisfaction.
The drawback of a customer-based decisioning engine is its need for accurate and efficient modeling systems and the requirement to have access in real time to multiple data systems. Many organisations wishing to use such a solution may lack the infrastructure or expertise in-house and therefore may also need to look externally for help with build and roll-out.
Choosing the right real-time decisioning engine
As each of these three technologies has its own set of benefits and challenges, an organisation must always consider its long-term strategic goals when deciding which option to implement. No single solution can be seen as a one-size-fits-all answer; however, with a sensible approach to people, process and desired outcomes, a practical solution that fits the brand, respects the customer and achieves a balance between short-term performance and long-term business sustainability can be achieved.
Luke McKeever
CEO, Portrait Software