Stop Optimizing for Your Existing Portfolio
The auto loan campaign looked bulletproof on paper. A $480 million credit union in the Midwest had built its prescreen criteria around the profile of its best-performing borrowers: members with 720+ credit scores, eight-plus years of tenure, and a history of direct deposit. Response rates from this “golden” segment had consistently exceeded 2%.
So when leadership decided to grow the portfolio by 15%, they did what seemed logical: they pulled a prescreen list of non-members who matched this same profile. Six weeks later, the results were dismal—response rates below 0.1%, cost per funded loan way off the benchmark, and a pipeline filled with applicants who ghosted after the first rate quote.
What went wrong? The credit union had fallen into a trap that new research from MIT Sloan Management Review identifies as one of the most common—and most costly—mistakes companies make when scaling into new markets.[1]
The Familiar-Market Fallacy
According to research analyzing over 1,000 technology companies, executives face a critical choice when expanding: learn from “familiar” customers whose feedback is easy to interpret, or learn from “target-market” customers whose preferences actually match the audience you’re trying to reach.[1]
The study found that familiar users provide clearer signals—their behavior patterns make intuitive sense to leadership. But target-market users provide more transferable signals, meaning their preferences align with the broader population you want to serve long-term.[1]
For community banks and credit unions running prescreen campaigns, this distinction is everything. Your existing high-performers—those loyal members with pristine payment histories—are familiar. You understand their rate sensitivity, their channel preferences, their likelihood to respond to direct mail versus digital. But when you use their characteristics to model conquest audiences, you’re assuming that non-members who look similar on paper will behave the same way.
They often don’t.
Why Lookalikes Mislead in Credit Marketing
Consider what makes your best current borrowers “best.” Often, it’s not just their credit profile—it’s their relationship with your institution. They’ve banked with you for years. They trust your brand. They respond to your offers because they already know you, not because the offer itself was irresistible.
Non-members with identical credit characteristics lack that relationship equity. They’re evaluating your offer against competitors they may already trust more. The variables that predict success within your portfolio—tenure, product depth, engagement history—simply don’t exist for prospects in the prescreen universe.
The MIT research puts it bluntly: “A product that resonates with users in one environment may fall flat in another, even when the differences seem minor on the surface.”[1]
This explains why so many prescreen campaigns underperform. Financial institutions optimize offers based on what works for existing members, then express surprise when those same offers fail to move the needle with conquest audiences.
When to Learn From Your Portfolio—And When to Look Outside
The research identifies two factors that determine whether familiar-market learning helps or hurts: how similar customer preferences are across markets, and how diverse your familiar market is.[1]
For prescreen marketers, this translates to practical guidance:
- If your current membership skews heavily toward one demographic or credit tier, their behavior patterns are unlikely to transfer to different segments. A credit union with 70% boomer membership will learn little about millennial acquisition by studying existing member response rates.
- If you’re targeting a segment already well-represented in your portfolio, familiar-market insights become more valuable. Expanding auto lending to a new county where demographics mirror your current footprint is different from launching a first-time homebuyer program when your average borrower is 54.
- If early campaign responses from target segments diverge sharply from portfolio benchmarks, trust the new data. Those signals—however counterintuitive—reflect how your actual prospects behave, not how you expect them to behave.
Building Prescreen Strategies Around Target-Market Learning
The most sophisticated prescreen programs treat early campaign waves as learning opportunities, not just production runs. They deliberately test offer variations across target segments before scaling, using response data from actual prospects—not portfolio proxies—to refine criteria.
This means accepting that your initial assumptions may be wrong. The rate that converts your existing members may not convert prospects. The channel that drives response internally may underperform externally. The credit tier you’ve historically avoided may represent your biggest growth opportunity.
Bureau data makes this kind of target-market learning possible at scale. Rather than guessing which prospect characteristics matter, institutions can test hypotheses against actual prescreen response patterns, iterating toward offer structures that resonate with the consumers they’re actually trying to acquire.
The Community FI Advantage
Large banks often default to familiar-market strategies because their sheer scale makes portfolio-based modeling statistically defensible. Community banks and credit unions can’t afford that luxury—but they also don’t need it.
Smaller institutions can move faster, test more aggressively, and adapt offer strategies based on real-time target-market feedback. That agility becomes a competitive weapon when combined with AI-powered prescreen technology that surfaces which prospect segments respond to which offers—regardless of whether those patterns match existing member behavior.
The institutions winning the prescreen game aren’t the ones with the biggest portfolios to model from. They’re the ones willing to let target-market data overrule familiar-market assumptions.
Your best borrowers taught you how to serve them. Your next borrowers will teach you something different—if you’re willing to listen.
