This shipping company needed to increase the number of its trade finance customers by more than 500% by 2017. To drive this exponential growth, the trade finance team realized that it needed to modernize how it markets to its 1 million customers, this would involve moving from proactive, highly targeted marketing to “ideal” sub-segments of customers.
The challenge would be identifying the ideal customer segments for each marketing campaign. The company has over a million customers, and traditional heuristics such as “customers who have the highest revenue” don’t work for trade finance targeting, as those large customers typically don’t need financing.
Complicating matters further, the company’s existing, manual segmentation process was slow, tedious and unable to scale across its 1,000,000+ customer base. Nor could it incorporate the features like seasonality, trade lanes, and networked presence, which the company wanted to use to target customers and determine, for example, what types of campaigns were best suited to customers with certain attributes. Other limiting factors included slow, manual data feeds and undocumented business semantics; lack of data integration to different systems; poor data quality; lack of extensibility to support business growth or a 360 view of customers; and limited capabilities to mine and analyze data. Furthermore, with its existing systems, the decision-making process was manual, not reproducible, and dependent on end user knowledge and expertise. This made it difficult to, for example, compare all available information on several customers at once and visualize the results for quick decision making.
Given these issues, management needed a new, accurate, and automated way to create pre-qualified leads and identify sales opportunities based on seasonality, trade lanes, and networked presence. The solution needed to:
- Provide access to all relevant information for the list of customers identified as the “best leads” to approach for financing
- Make the decision-making process faster and easier
- Automate the process of generating leads and ranking them
How Maana Helped
Maana created a new knowledge application and deployed it external to the company’s cloudinstance. To create it, Maana’s team took a snapshot of structured data from the company’s database (for example, Account Receivable, Booking, Dispute, and Write-off data) and modeled it using the Maana Knowledge Platform.
This model used a ranking system based on the “Customer Trust Score” that the company assigned to each customer; the higher the score, the more “attractive” the customer would be from targeting perspective. Maana worked with the Trade Finance team to identify all relevant features that should contribute to the final score for each customer, which would ultimately be used to identify the best customer leads for financing. Each feature was calculated from the raw data using one or more tables for calculating each feature; based on the resulting value, each feature was ranked and assigned a score. The Trust Score is the sum of all scores for features for a particular customer.
Now, the company’s salespeople, data scientists, and marketing analysts can make better, faster decisions regarding customer segmentation and campaign targeting, as the solution gives them:
- All relevant information about each customer at a glance
- A pre-ranked list of customers by score (the higher the score, the more “attractive” the customer to approach for financing)
- The ability to visually compare the most important features for multiple customers at once
- Insights on customer seasonality behavior
Armed with this data and insights, the trade finance team can accurately identify which customers to pursue in minutes, rather than hours. Equally important, they can identify ideal small and midsize customers as targets; this will be key scaling the solution with partners.