AI Research Applications
AI Research is currently working on several cutting-edge analytic products:
Optimal Assignment Engine
Many marketers already deploy sophisticated probabilistic models to select the best targets for a variety of promotional activities. Yet, there are situations when even model-based scores are not sufficient tools to generate maximum ROI. Marketers offering multiple products with overlapping customer eligibility criteria and other constraints (contact rules, budget, etc.) are familiar with such challenges and have often ad-hoc and sub-optimal solutions. There exists an optimal solution for such combinatorial assignment problems, which will maximize overall profit, across multiple promotions. Read more . . .
Personalization Analytics
Traditional direct marketing strategies have reached a high level of analytical sophistication. However, most strategies are only able to address the product side of the ‘Customer - Product’ equation. Particularly in today’s online marketplace, such product-centric approaches are neither sufficient, nor always feasible. Personalization Analytics is a rapidly emerging discipline, which combines computational advances with sophisticated analytics, to deliver a customer-centric marketing strategy. It aims to deliver tools that will enable marketers to drive top-line revenue by capitalizing on the expressed interests of their customers. Read More . . .
Transaction Data Analytics
Most data-savvy marketers are already aware of the fact that invaluable knowledge about their customers is trapped within the data housed in their data warehouses. The ability to discover this knowledge is essential to succeed in the race to capture the most of their customer’s spending. Not all customer data are alike, i.e. the quality of the ‘secrets’ they hold vary. Customer transaction data surpass all other data (personal finance, account application, billing, demographic, census, etc.) in how much they can reveal about a customer’s spending behavior; yet, this information is not easily accessible by traditional extraction methods. Read more . . .
