Using Data to Drive Credit Card Spend



A FinTech client had the objective to become the first card in the wallet for its credit card customers, but it had lost faith in spend stimulation promotions.  In the past, they saw a lift in sales for the duration of the promotion, and immediately after, the spend decreased to its previous levels.  Additionally, the ROI tended to be negative because:


  1. Promotions tended to look as follows: spend at least $X and get Y points, and
  2. Setting the level of the X was incredibly difficult. For many cardholders, $X was unattainably high, and they wouldn’t even bother trying.  And for another large population, $X was too low, and they would have spent that much in any case.  This means that whatever incentive the client was to give out would be too much and have a negative ROI.


What was our solution?  It was 2-fold:


  1. Create a calendar of promotions. That way, cardholders would be rewarded at irregular, but frequent intervals, and get into the habit of using their credit card for everyday purchases.
  2. Set the $X threshold at the individual, personalized level.


Personalized promotions allow you to set relevant goals for your customers: not too high or low, but just right to drive the desired behaviour.


The first step was easier, but how did we accomplish the 2nd


  • To establish $X on a customer-level basis:
    1. We leveraged external data to estimate how much cardholders spent on competitor credit cards and compared it to how much the cardholder spent during the same time last year. For example, Cardholder A spent $300 with our client and $1,000 with their competitor for the month of April.
    2. We made assumptions on how much share of the cardholder’s spend our client could conservatively bring over from the competitor for the duration of the program. For the cardholder above, we could assume that we could bring 10% over from the competitor (i.e., 10% of $1,000 = $100).
    3. We set the $X threshold at last year’s spend plus the share they would bring from their other card. For Cardholder A, that would mean $300 + $100 = $400.
  • Then, we went one step further. How much should we reward customers for their increased spend, share of wallet, and loyalty?  This was based on three factors:
    1. How much the client could afford set the parameters for each offer.
    2. Customer lifetime value. The best and next best cardholders were rewarded better than the rest.
    3. How much of a stretch we were asking of each cardholder. This means the more we asked someone to increase their spend over their previous year’s spend, the higher the compensation.

This created a one-to-one digital communication, as each recipient’s threshold and reward combination was unique to their circumstances.  It was relevant, personalized, and went well beyond knowing their first name. 


This ended up being the first of its kind promotional campaign for our client and a hugely successful one at that. 


This is just one example of the power data can unleash when used properly.

More to Explore