Entering a new B2B segment? Here’s how to optimize approval rates while reducing risk

The learning curve can be steep when entering a new market segment: How long is the sales cycle? How loyal are customers to their existing suppliers? And—most importantly—does your existing credit model address the specific risks of this new segment without discouraging new sales?

Time and test marketing will provide answers to the first two questions. When it comes to credit, however, reaching a solution can be much quicker—especially if you create a customized credit model with instant decisioning and ongoing monitoring.

Why automated risk decisioning matters

Companies often make case-by-case credit decisions based on applicants’ bureau commercial credit files and individual credit analysts’ preferences. Each analyst has their own bias on what they like and don’t. Sales pressure, especially on big accounts, can also play a role. In addition to eating up critical staff time, this approach produces inconsistent results—and may be especially dangerous when entering a market that’s unfamiliar to your credit team.

An automated decision engine will provide you with real-time risk assessment and a credit decision, with an immediately available credit line.  You’ll find that you can likely provide instant decisions for the majority of applicants and significantly reduce the cost of judgmental credit review, since only marginal applicants or large line requests get flagged for manual review.

We’ve seen companies achieve between 80-90% of instant decisions from all incoming online applications. The rest have required some level of manual review.  This automation allows you to remove all manual time spent on making routine decisions and spent making routine approvals and declines.

Adapt your existing credit model

When entering a new market segment, it’s critical to update your credit model so it reflects this market’s risk levels and growth goals. You’ll want your models to score applicants based on bureau data as well as variables that are statistically significant predictors of severe delinquency or charge-offs.

Ideally, you’ll want to use a combination of historical bureau data and internal performance review to evaluate customers; however, this approach can be more challenging with newer segments. If you’re venturing into untested markets, ask a credit bureau for a sample data set to use as a starting point on which to build your scoring model, and always opt for real-time integration from bureaus.

A new market segment may require a new credit model altogether. For example, if the average collection period for a particular segment runs on a longer schedule than what you’re used to servicing—say, 90 days versus 45 days—you may consider building a separate model that incorporates variables that are more predictive of credit risk.

Frequent credit monitoring

Once a customer’s credit is approved, it’s easy—especially in a strong economy—to coast along assuming their risk profile isn’t changing much. But our data shows that consistency is not the rule when it comes to commercial credit.

Consider the results of our analysis: When we re-scored more than 2,000 customers after a year, only 17% had no change to their risk profile. All of the others – 83% – saw their risk profile change. That’s a surprising lack of stability.

Within this group, some customers’ risk profiles improved—which means they may be good candidates for a larger credit line that offers more purchasing power. Of course, others’ risk profiles worsened.

As time goes on, compare the behavior of your new customers with the sample data set, and adjust your model accordingly. By re-validating your models regularly, you can ensure that risk criteria aren’t too strict—or too loose. This way, the models don’t turn away prospects with good credit, or offer it to those who present a real risk of delinquency or losses.

The objectivity of an automated model insulates your decisions from errors caused by human perception: While a credit manager might not flag a previously pristine customer that recently delayed payment, the behavioral model will.

The upshot: A customized, automated credit model can reduce risk and boost approvals

As you focus on growth in a particular market segment, consider an automated credit model with instant decisioning, a tailored scoring model and ongoing credit monitoring. Our experience shows that up to 90% of online applications could be decisioned this way, and ongoing monitoring can protect you from the frequent changes in risk profiles. These solutions can help ensure that you’re taking only the appropriate risks as you expand your business.

Written by Sarah Faatz, Director of Credit Risk Management

 


Interested in learning more?
Check out Why a 24-hour Credit Approval Isn’t Fast Enough