The Transformative Power of Marketing Decision Engines

By Lakshmana Gnanapragasam, SVP — Strategy & Insights, Epsilon India & APAC

In the past decade, there has been a steady shift across enterprises in how they make business decisions — from traditional analytics methods to an automated and highly scalable way of analytics called ‘decision engines.’

In the traditional analytics setup, enterprises often build specialized teams to solve for customer data and analytics roadmaps; analytics platforms/infrastructure; strategic customer insights; advanced customer and marketing analytics; and measurement and attribution.

Having realized the importance of each of these teams and their components, enterprises look for proven expertise in solving these challenges in their teams. While this approach had worked in the past, it’s also time-consuming for enterprises to go from data to realized business benefits. This is due to the involvement of multiple teams and hand-off points that create friction in maximizing the full business benefits from the use cases.

This traditional approach is being challenged by ‘Marketing Decision Engines’ that attempt to collapse the multiple components into one solution, by improving an enterprise’s ability to capture maximal value from use cases and helping accelerate the time to value realization.

Let’s consider the use case of running retention campaigns on an existing customer base for a multi-channel retailer. This would require bringing disparate customer data such as demographics, purchase behaviour, promotions history, loyalty transactions and online behaviour together, and scoring the existing customer base for ‘attrition probability’.

These customer level attrition scores are then fed into a marketing orchestration tool to select the list of customers for campaigns on owned channels such as email, SMS, mobile/web push, etc. Now, the marketing orchestration tool takes over, executes the campaigns and produces the results. These results would be further analysed by a team to show how the campaign has performed and the business value delivered.

This approach, while has worked in the past, is sluggish and not scalable due to the proliferation of the customer data, complexity of marketing tools and technologies and drastic changes in customer behaviour and journeys.

To solve these challenges, enterprises are approaching this as a unified decision engine solution. And a good decision engine solution has six critical components.

1. Decision engine strategy: This requires some upfront thinking to determine what the enterprise wants to achieve through its engines. This might involve discovering data, identifying top scenarios, determining business cases and developing roadmaps to achieve the outcomes.

2. Infrastructure: This refers to the analytics and decisioning environment with the necessary data and tools available for analysts, data scientists and data engineers to work on.

3. Data integrations: This brings relevant data from across the enterprise into the decisioning environment and requires batch and/or real-time integration with various internal and external data sources.

4. Marketing operations: Front end interface marketing analysts can use this to perform their role. This might include audience selection, campaign design, rules set-up, choice management, test set-up, and customer journey orchestration.

5. Analytics and algorithms: Developing machine learning models, maintaining model repository, deployment of analytical models and continuous tuning and enhancements of the models.

6. Integration services: Integrating the decision engine environment with various marketing channels (owned and paid) for batch and real-time to deliver the marketing use cases.

Now, let’s go back to our retention use case to see how this decision engine solution provides wings to marketers to do more, and do it better. This fully automates the process of ingesting, preparing, and scoring the customer data for customer level attrition scores. This data is available for the marketer within the decision engine environment. The marketer can take full control of their campaign by using the front-end interface to set-up their campaign and execute it. Results are available as soon as practical for the marketer to evaluate the campaign and decide on the next set of actions. This reduces the time to market for campaigns and improves the odds of achieving the desired business outcomes.

With this solution in place, the marketer can now think beyond the current campaign. For example, data integration with the brand’s website and app would help the marketing team to identify customer triggered events such as browsing a product page, adding to a shopping cart, etc. and take immediate actions through outbound campaigns. Through an ad-tech integration, there are enormous opportunities to identify customers and prospects with similar online interests and behaviours as that of existing high value customers, and to bring them to the brand’s online properties and offline stores.

A few business results we’ve seen with decision engines so far include:

- 52% increase in the online conversion rates driven by a product and offer personalization engine for a multi-channel beauty retailer

- 90% improvement in the cycle time associated with setting up and running outbound marketing campaigns for an automobile retailer

- 200% lift in the engagement rate through content affinity and personalization models for a media organization

Marketers don’t need to be overwhelmed with the seeming complexity of the decision engine solution. All components don’t need to be in place on day 1. The stand-out enterprises took on an experiment and learnt to approach this by thoughtfully evaluating the use cases and infrastructure requirements and made investments in their decision engine foundations. And they continue to build on top of that foundation as more use cases open-up for them.

If you haven’t thought about implementing marketing decision engines yet, the time is now.

Also published on The Times of India: https://timesofindia.indiatimes.com/blogs/voices/transformative-power-of-marketing-decision-engines/