We invite you to review our real-world stories about how we helped match the offerings of our clients with the unique demands of their customers.


Additional details on these case studies, as well as client testimonials, are available upon request.

How to Implement a Pre-approved Lending Program

Challenge

In the midst of market instability, a national banking client was struggling to maintain its ITA lending programs. Low response rates, and even lower approval rates, were resulting in per-account acquisition costs that were unacceptably high.

Solution

We worked directly with our client to revise the lending strategy, and implemented a custom pre-approved direct mail program to build a more cost-effective portfolio of high quality loans for home equity, mortgage, auto and credit card.

How We Delivered

  • Data-Driven Strategy

    We kicked-off an analysis of application data, looking at characteristics of approvals and declinations, and conducted a thorough review of product level credit risk guidelines. Based on this work, we were able to direct our credit bureau partners to set up custom pre-approved criteria in alignment with credit risk policy.

  • Technical Expertise

    In order to develop a learning loop data environment, we requested extracts from different legacy systems. We also set up a hosted datamart, which allowed us to automate reporting via a secure web portal. And our archives of snap-shotted predictive data helped our client set the stage for future modeling.

  • Nimble Responsiveness

    We worked closely with our client and longtime partners to address the diversity of legal, risk, compliance, product and operational issues while cutting through the red tape.

How to Enable Rapid Prototyping

Challenge

With internal resources stretched thin, a large regional bank lacked insight into the success of their direct-mail campaigns and wanted to take advantage of untapped internal data stores.

Solution

As a trusted extension of our client's internal team, we brought in our own partners to help review the current marketing infrastructure. Carderock was then able to introduce an innovative step in the process to securely exchange key data with the large existing vendor and kick-start targeting efforts. The outcome was a win-win with that vendor who was able to implement change in a measured pace, as well as for our client, who could cost-effectively test a variety of programs.

How We Delivered

  • Data-Driven Strategy

    As with every campaign, we constantly look to incorporate multiple models to develop various audience tests. Our Learning Loop environment allows us to not only review high level result totals, but to also study the underlying characteristics that drive behavior.

  • Technical Expertise

    From a regularly scheduled blinded extract, we applied a series of model scores to determine which product was the best for each individual in the household. We then triaged with our client to assign the final offer.

  • Nimble Responsiveness

    Our client wanted to test an internal trigger program, which typically would take at least 3 to 4 weeks of setup. Instead, we developed a programming solution that could be accomplished in days without bypassing numerous quality control steps.

How to Uncover Surprises through Modeling

Challenge

A leading direct-to-consumer mutual fund company suspected that an inordinate amount of assets were concentrated among a small pool of customers. The pool needed to be enlarged either through new customers with hidden wealth, or among those with a high potential for generating wealth.

Solution

With a focus on small business owners already targeted, we created a custom predictive score. The findings from our model, however, came as a surprise to us all. It led the team to develop a completely different marketing strategy, which resulted in a new allocation of marketing dollars.

How We Delivered

  • Data-Driven Strategy

    We reviewed the modeling process and revisited project objectives and assumptions to ensure that each variable made logical sense. While the findings did make sense, they were not expected, and drove a change in our client's marketing strategy.

  • Technical Expertise

    We looked at not just demographics or profiles, but time series data with regard to product ownership, tenure, channel usage, contact history, and transaction behavior as well as existing model scores. We also reached out to external vendors to find additional predictors of investable assets.

  • Nimble Responsiveness

    With a tight 6-week window, we met with internal groups, reviewed warehouse data models, and prioritized desired data elements. We then developed an effective sampling strategy to reduce the volume of data to be extracted, while meeting our project objectives.