Print | Share

A Winning Strategy

September 2017: Vol 40 No 9
Best practices for developing a data-driven credit marketing plan

Marketing analytics data hovering over a tablet PCSponsored by Experian

Many credit unions take a leap of faith when it comes to developing prospecting strategies. But effective marketing strategies are developed from deep analysis with clearly identified objectives. They are constantly evolving—no setting and forgetting. So, what are the basics of optimizing your prospecting efforts?
 

Establish Goals


Unfortunately, far too many discussions begin with establishing targeting criteria before program goals are set. This leads to confusion.
Developing targeting criteria is kind of like squeezing a balloon—when you restrict one end, the other tends to expand. Imagine the effect of maximizing response rates when soliciting new loans. If no other criteria are considered, you could end up targeting high-risk individuals who cannot get approved elsewhere.

Obviously, we’re not interested in increasing originations at all cost; risk must be understood as well. But this is where things get complicated. Lower-risk consumers tend to be the most coveted, get the best offers and therefore have lower response rates and margins.
 

Simplicity Is Best  

    
In the 1960s the U.S. Navy developed the KISS acronym (“keep it simple, stupid”) on the philosophy that complexity increases the probability of error. This is largely true in targeting methodologies, but don’t mistake limiting complexity for simplicity.

Perhaps the most simplistic approach to prescreen credit marketing is using only risk criteria to set an eligible population. Breaking a problem down to this single dimension generally results in low response rates and wasted budget. Propensity models and estimated interest rates are great tools for identifying consumers who are more likely to respond. Adding them as an additional filter to a credit-qualified population can help increase response rates.

But what about ability to pay? So far, we’ve considered propensity to open and risk, the latter being based on current financial obligations. Imagine a consumer with on-time payment behavior and a solid credit score who takes a loan, only to be unable to meet the obligation. You certainly don’t want to offer debt that will cause a consumer to be overextended. Instead of going through costly income verification, income estimation models can assist with identifying the ability to repay the loan you are marketing.

Simplicity is great, but not to the point of being one-dimensional.
 

Take off the Blindfold


Even in the days of smartphones and GPS navigation, most people develop a plan before setting off on a road trip. In the case of credit marketing, this means running an account review or archive analysis.

Remember that last prescreen campaign you ran? What could you have accomplished with a more sophisticated targeting strategy? Having archive data appended to a past marketing campaign allows for “what if” retrospective analysis. What could response rates have been with a propensity tool? Could declines due to insufficient income have been reduced by leveraging income estimation?

Archive data gives 20/20 hindsight on what could have been. Just like consulting a map to determine the shortest distance to a destination or the most scenic route, retrospective analysis on past campaigns allows for proactive planning for future efforts.
 

Practice Makes Perfect


Even with a plan, you probably still want to have the GPS running. Traffic could block your planned route or an unforeseen detour could divert you to a new path. Targeting strategies must continually be refined and monitored for changes in customer behavior.

Test and control groups are essential to continued improvement of your targeting strategies. Every campaign should be analyzed against the goals and KPIs should be established at the start of the process. New hypotheses can be evaluated through test populations or small groups designed to identify new opportunities. 

Let’s say you typically target consumers in a risk range of 650 to 720, but an analyst spots an opportunity where consumers with a range of 625 to 649 with no delinquencies in the past 12 months performs nearly at the rate of the current population. A small test group could be included in the next campaign and studied to see if it should be expanded in future campaigns.
 

Never ‘Place Bets’


Assumptions are only valid when they are put to the test. Never dive into a strategy without testing your hypothesis. The final step in implementing a targeting strategy should be the easiest. If goals are clearly understood and prioritized, past campaigns are analyzed, and hypotheses are laid out with test and control groups, the targeting criteria should be obvious to everyone.

Unfortunately, the conversation usually starts at this phase, which is akin to placing bets at the track. Ever notice that score breaks are discussed in round numbers? Consider the example of the 650 to 720 range. Why 650 and not 649 or 651? Without a test-and-learn methodology, targeting criteria ends up based on conventional wisdom—or worse, a guess.

As you approach strategic planning season, make sure you run down these steps in this order to ensure success next year.
1.    Establish program goals and KPIs.
2.    Balance simplicity with effectiveness.
3.    Have a plan before you start.
4.    Begin with an archive.
5.    Learn and optimize.
6.    In God we trust; all others must bring data.

Experian serves credit unions across all asset tiers with a team 100 percent dedicated to this industry. Beyond credit reports, the company offers data analytics, segmentation solutions, automated decisioning and digital strategies that help credit unions drive growth, compete through differentiation and protect against fraud and regulatory risk.