Redefining ‘Character’: A Better Approach to Relationships in Lending Decisions

Image details: Historic Edmund Pettus Bridge, Selma, Alabama. Site of the 1965 Selma to Montgomery marches that transformed the fight for voting rights.

This post is part of a series examining how banks can reimagine risk, based on the latest insights from Underwriting for Racial Justice.

By Erin Kilmer Neel, Executive Director of Beneficial State Foundation 

There is a growing debate in the banking industry about credit scores, which often fail to fairly measure a borrower’s likelihood to repay. To address this, credit bureaus are adding data like rent and cellphone payments, while fintechs and lenders are exploring alternative ways to assess “character,” using indicators such as cash flow, social media activity, or even cellphone battery usage. While these are creative and some are potentially promising, more data doesn’t always lead to more financially inclusive or accurate assessments, and some of the non-financial data can feel very invasive for an applicant. There are other thoughtful ways that lenders are looking at the likelihood of repayment as well. 

As part of Beneficial State Foundation’s Underwriting for Racial Justice (URJ) program, we’ve explored another way lenders assess “character” beyond credit scores–through borrower relationships and interactions. Almost all lenders–from CDFIs to large banks–use relationships and applicant behaviors to assess likelihood to repay, often without explicitly naming, structuring, or evaluating these practices. How this information is used can either help or hurt borrowers in unintentional, and sometimes inequitable, ways.

Read on to explore the real-world patterns we’ve observed, uncover unconscious practices, and get tools to sharpen your approach to assessing “character” in ways that are more intentional, transparent, and equitable.

What Is Relationship- or Character-Based Lending?

Relationship-based lending (also called character-based lending) supplements or replaces credit scores and credit reports with contextual information gleaned through relationships—whether directly with the applicant, their community, or the lending institution’s partners. These methods are especially common among Community Development Financial Institutions (CDFIs) and community-based lenders, particularly to account for the fact that credit report approaches can unfairly disadvantage people of color due to longstanding economic inequities.

Examples of lenders intentionally using character-based lending include Denkyem Coop, Native Women Lead, Boston Impact Initiative, B:Side Fund, Mountain BizWorks, Runway, and Kiva. Their approaches often incorporate practices such as gathering personal references, fostering mentorships, using community-based vetting, and obtaining referrals from trusted technical assistance (TA) partners.

Looking at lenders like these is insightful because they are intentional about utilizing relationships and behaviors as part of the decision-making process. Many lenders draw on similar information, but often not consciously or systematically, as part of their credit assessment.

Conscious and systematic methods might include:

• Gathering references.
• Engaging in incubator programs that build relationships between the lender and potential borrower well before they apply for a loan.
• Asking borrowers about how they plan to repay if they encounter financial difficulties.
  

Unconscious methods might look like:

• A board member vouches for an applicant: “She’s solid – we go way back.”
• A loan officer fast-tracks an application because the applicant is a family friend.
• An underwriter gives a borrower the benefit of the doubt because “they’re a regular at our church.”

Recognizing that relationships influence lending decisions is the first step toward making these practices more intentional, transparent, and fair.

When left unexamined, these unconscious relationship biases can lead underwriters to move  faster or slower, apply more or less due diligence, or unintentionally reinforce preconceptions about an applicant with the information they gather. Recognizing that relationships influence lending decisions is the first step toward making these practices more intentional, transparent, and fair.

How Relationships Show Up in Practice

Whether through direct relationships, close connections, partners, neighborhoods, alliances, shared identities, or clout, the nature of an applicant’s relationship to the decision-maker can impact the loan decision in many ways. Often, these relationships aren’t formally part of underwriting criteria–but they still shape how an applicant is treated—sometimes subtly, sometimes significantly.

Let’s look at a few dynamics that have surfaced during our work. These are example moments where people realize their existing practices aren’t neutral and may need rethinking.

Scenario 1: An applicant with a thin credit file applies for a small loan. She has banked with this institution for years, maintains a steady cash flow, and has never overdrawn her account. But because she isn’t personally known to the loan officer, her application is judged strictly by her credit score. With no one stepping in to provide context, she is denied – even though her history with the bank suggests she could likely repay.

Scenario 2: That afternoon, a small business owner who often brings lunch from her restaurant to the bank staff, comes in with some sandwiches and cookies. She decides it’s time to apply for a business loan, the loan officer actively advocates for her, telling the underwriter, “She works so hard and is always so kind to us. Let’s see what we can do to make this happen.”

These two scenarios show how relationship-based lending can cut both ways. In one case, lack of personal connection left a qualified applicant without access to credit. In the other, familiarity gave someone an advantage that others might not receive. Both outcomes warn of inconsistency: lending decisions influenced too heavily by personal ties can open the door to favoritism and bias, while decisions without any relationship context can overlook good borrowers.

While the stories are not necessarily examples of wrongdoing, they show how relationship-based lending often happens quietly, without structure or oversight. This can privilege applicants with more social capital, often white, native-born, or well-connected individuals, while disadvantaging others who have a kind of “relationship thin file.”

The equity implications are significant. Because of redlining, segregation, and bank disinvestment, people of color are less likely to have ties with bank staff, board members, or social networks that create these informal advantages. When personal familiarity drives lending decisions, it risks deepening inequities rather than reducing them.

Yet the same dynamics hold potential. With structure and intention, relationship-based practices can strengthen fairness and improve underwriting. For example, referrals from trusted technical assistance providers or vendors, especially in communities of color, can add valuable context about an applicant’s reliability and business behavior.

The key is ensuring these practices are visible, standardized, and equitable so they benefit everyone, not just the well-connected.

Designing for Fairness: Key Areas to Rethink

The point is not to abandon character-based assessments but to apply them thoughtfully and equitably.

Lenders often have a clear sense of the traits they value when assessing an applicant’s “character.” The challenge is not identifying these traits, but recognizing that the ways they are evaluated can unintentionally disadvantage applicants of color. For example, expecting perfectly completed applications may penalize those with less formal education or limited English proficiency. Requiring references from “respected” community members can exclude those whose networks do not overlap with the lender’s. Even seemingly neutral expectations, such as proactively asking for help or presenting detailed backup repayment plans, may be more accessible to applicants who have had prior exposure to loans, formal business training, or multigenerational family support.

The point is not to abandon character-based assessments but to apply them thoughtfully and equitably. With intentional processes, lenders can identify which criteria expand access and which reinforce inequities, then structure practices to ensure consistency and fairness. Done well, character assessments can become tools that help level the playing field rather than deepen existing divides. 

To make character-based lending work better and more fairly, lenders can reflect on three critical areas. A short list is provided below, but for a more detailed assessment, please utilize our free downloadable Character and Relationship Assessment Review Tool.

1. The Information You Use

Ask yourself: 

• • What are we really trying to learn about the applicant – trustworthiness, preparation, or financial behavior? 
  • Are we relying on signals that are truly predictive of repayment, or just comfortable ones that may reflect privilege (like polished communication)?

Recommended steps:

• • Use your own portfolio data to identify which signals actually predict repayment. If unavailable, seek out market data.
  • Standardize questions that help applicants explain their plans and showcase their resilience, especially for thin credit files.
  • Remove application requirements that do not relate to repayment ability to reduce burdens on applicants and staff.

2. The Relationships You Rely On

Ask yourself: 

• • Which relationships influence our decisions, and are they equitably accessible?
  • Do the relationships provide relevant repayment insights, or do they risk favoritism?

Recommended steps:

• • Require documentation in credit memos if relationships informed a decision, and how they were validated.
  • Diversify TA partners and referrers, and consider community advisory input.
  • If using outside information (such as social media or reviews), be transparent and allow applicants the opportunity to explain context.

3. The Processes You Use

Ask yourself: 

• • Are our applications, interviews, and follow-ups designed with transparency and equity? 
  • Do our processes unintentionally penalize those with multiple jobs, limited English, or lower access to technology?

Recommended steps:

• • Simplify forms, provide translations, and offer reference templates for vendors or landlords.
  • Normalize applicants asking for help by creating a culture where questions show engagement, not weakness.
  • Offer flexible ways for applicants to share their story, such as presenting directly to credit committees.

Character-based and relationship-based decision-making can either open doors or reinforce barriers, directly influencing who gets approved for credit. Every lender has a choice: to adopt intentional practices that expand opportunity, or to let unconscious habits quietly limit it.

Done well, character assessments can become tools that help level the playing field rather than deepen existing divides.

Tools and Templates

The Character and Relationship Assessment Review Tool and a Sample Policy can be downloaded and adapted to fit your institution. We encourage you to review the sample policy, edit it for your own needs, and reach out if you would like assistance tailoring these practices to your context.

• Character and Relationship Assessment Review Tool 
• Sample Lender Policy on Relationships in Lending 

By making character and relationship-based practices visible, standardized, and equitable, lenders can turn what has too often been a source of bias into a powerful driver of fairness, trust, and expanded opportunity.