AI excels at identifying creditworthy prospects from bureau data, but knowing when not to automate is a strategic competency. Learn the 80/20 framework for balancing automation with human judgment in prescreen campaigns.
New research shows 89% of bank clients say offerings lack personalization. Community FIs that master data-driven prescreen marketing can capture borrowers fleeing impersonal experiences.
Credit unions are absorbing more credit risk than banks—by design, not distress. Learn how to convert this counter-cyclical strategy into funded loans through proactive borrower acquisition.
The most valuable thing prescreen campaign analytics can do is tell you where you’re winning and where you’re wasting money. A review of aggregated weighted campaign performance data across offer types and campaign audiences reveals a clear pattern — one that holds up under a consistent evidence standard, and contains one finding that should change how credit unions think about member acquisition.
Market share is defined as your institution’s closed loans in a given product category divided by the total loans closed in that category across all competing institutions serving the same population that received your firm offer of credit. Conversion rate measures what your campaign did in isolation. The two can move in opposite directions — and when they do, conversion rate is wrong.
| Campaign 1 | Campaign 2 | |
|---|---|---|
| Offers Sent | 1,000 | 1,000 |
| Your Loans Closed | 20 | 25 |
| Total Market Closings | 200 | 500 |
| Campaign 1 | Campaign 2 | |
|---|---|---|
| Offers Sent | 800 | 600 |
| Your Loans Closed | 24 | 15 |
| Total Market Closings | 120 | 50 |
Most institutions evaluate campaign performance with conversion rate: the number of members or prospects who responded to an offer divided by the number of offers sent. It’s intuitive, easy to compute, and almost entirely useless as a measure of competitive improvement.
Conversion rate measures what your campaign did. Market share — defined here as your institution’s closed loans in a given product category divided by the total loans closed in that category across all competing institutions serving the same population that received your firm offer of credit — measures how you performed relative to everyone else competing for the same borrowers at the same time. Institutional competitiveness is a relative concept, not an absolute one. A credit union’s goal is not to convert more offers; it is to win a larger share of the lending market it operates in. Market share is the only metric that directly measures that.
The distinction matters because conversion rate and market share can move in opposite directions — and when they do, conversion rate produces a false signal as described in the illustration above.
All findings in this analysis are expressed in market share terms. Conversion rate appears only where it is arithmetically necessary to derive market share from reported data.
To separate reliable signal from noise, findings here require a minimum of 24 total sales — target and other product combined — on both the existing and prospective side before any comparison is treated as actionable. That threshold eliminates most of the data. Five bilateral comparisons survive it. A separate disclosure applies to mortgage new purchase, explained below.
Aggregated weighted campaign data across participating credit unions and community banks. Minimum evidence threshold: 24 total sales (target + other) on both existing and prospective sides. Five bilateral comparisons survive. Market share = institution’s closed loans ÷ all loans closed among the same prescreened population.
Across all campaigns in the dataset, existing member campaigns capture 25.7% weighted market share versus 1.6% for prospective campaigns — a 16.5x gap supported by more than 3,200 total sales in existing campaigns and more than 2,200 in prospective ones. Every product-level finding that follows sits within this frame.
Personal and consumer loans are the strongest finding in the dataset. Existing member PCL campaigns capture 56.1% market share — nearly 1,000 total sales — while prospective PCL campaigns capture 2.2% on more than 200 total sales. More than half of existing members who enter the personal loan market chose their credit union. Fewer than 1 in 40 prospects did. No other bilateral comparison is this well-supported.
Auto loan refinance is the highest-volume comparison available. With 563 total existing sales and 1,225 prospective, the 32.0% versus 2.0% gap is the most statistically reliable finding in the data. ALR is also the only prospective product with a large enough sample to model cost-per-acquisition with genuine confidence.
Auto loan extensions show the same structural pattern at 19.3% existing versus 1.1% prospective across 424 and 662 total sales respectively. The relationship advantage is real, reproducible, and not dependent on small-sample luck.
Home equity products confirm the same direction — 22.0% existing versus 1.9% prospective — though both sample sizes just clear the minimum at 33 and 42 total sales. The direction is credible; the specific figures warrant more caution than the loan products above.
Mortgage new purchase is the most strategically significant finding in the dataset, and the one that requires the most careful reading. Existing member MNP campaigns show 4.3% market share; prospective MNP campaigns show 3.2%. The gap is 1.3x — compared to 12–25x for every other validated product.
Full disclosure on classification: The MNP existing data has an unusual composition: 12 target product sales (mortgages) and 144 other product sales, for a total of 156. That 12:144 ratio of target-to-other is far outside the norm for other offer types and raises a legitimate question about whether some of those 144 sales involve mortgages that were attributed to a different product category. If any meaningful number of the 144 other-product sales were actually mortgages coded differently, the true existing market share could be materially higher than 4.3% — which would widen the gap and weaken the narrow-gap finding. The prospective side is less exposed to this concern, with 19 target sales and only 5 other product sales. This classification question is under separate review. The narrow gap is included here as a directional finding, not a settled conclusion.
External research suggests this is structural, not anecdotal. Independent mortgage companies now originate 63–72% of all home purchase loans (CFPB/HMDA, 2023–2024), meaning the majority of borrowers leave their primary financial institution to close a mortgage. Depositories’ share of mortgage originations has fallen from 81% to 39% over 15 years (Kansas City Fed). No other consumer lending product shows this pattern. Real estate agents influence lender choice for nearly half of home buyers, and agents recommend nonbank lenders twice as often as banks, citing speed and closing reliability (STRATMOR Group). The CFPB’s National Survey of Mortgage Borrowers found that 77% of borrowers applied with only one lender, but among those who did shop, the decision was almost entirely rate-driven—Freddie Mac found that rate dispersion across lenders doubled in 2022. Borrowers without an existing relationship have no reason to default to any institution; they are won on terms, timing, and offer. This is exactly the competitive dynamic that makes MNP different from PCL, ALR, ALE, and HCC, where the existing relationship produces a 12–25× structural moat.
Importantly, a mortgage win brings a new member inside the institution—where they enter the existing-member funnel and encounter the 12–25× campaigns going forward.
With that disclosure in view, the structural argument for MNP as an acquisition vehicle remains: home purchase is a life-event decision driven by rate and timing rather than relationship depth. Buyers comparison-shop regardless of where they bank, which should reduce the existing-member advantage relative to products where trust and convenience dominate. Whether the advantage narrows to 1.3x or somewhat more will be clearer as classification questions are resolved and as more campaign data accumulates.
Even with the classification caveat, the mortgage new purchase finding is worth acting on directionally. But the downstream logic deserves careful framing.
Oliver Wyman’s widely-cited analysis “Mortgage Cross-Sell: The Elusive Opportunity” found that for large banks, the mortgage-to-primary-relationship conversion is more assumed than achieved: primary banking relationships are sticky, most mortgage borrowers don’t consolidate their checking account with their lender, and only 10–15% of customers switch primary banks in any 18-month window. That finding applies to large commercial banks operating at arm’s length from their customers.
Credit unions are structurally different in three ways that change the arithmetic. First, credit unions predominantly portfolio their mortgage loans rather than selling them to the secondary market, which means the servicing relationship — monthly statements, escrow management, rate conversations — stays in-house for the life of the loan. Second, membership itself is a continuing engagement mechanism: once someone becomes a member to close a mortgage, they are already inside the cross-sell environment. Third, credit unions tend to service smaller geographic footprints where branch proximity and community identity reinforce the relationship in ways a national lender cannot replicate.
The hypothesis, then — supported directionally by this data but not yet proven within it — is that a prospective MNP campaign that wins a home purchase loan at or near existing-member conversion rates delivers downstream value no other prospective offer type can match. The new borrower becomes a member; the member encounters the same PCL, ALR, and HCC campaigns where existing members convert at 20–56% market share. Whether that flywheel actually closes, and at what rate, is a question this dataset cannot fully answer — but it is the right question for credit unions running MNP campaigns to measure going forward.
The rate environment reinforces the timing argument. High rates have suppressed purchase volume and reduced origination competition. Mortgages originated now become existing members for the next rate cycle — where this data consistently shows market share in the 20–56% range across personal loans, auto refinance, and home equity products.
For existing member campaigns, PCL, ALR, and ALE are the highest-confidence investments, supported by hundreds to thousands of total sales each. All three should run continuously at maximum eligible coverage. HCC belongs in the mix with appropriate precision caveats.
For prospective campaigns, MNP is the only offer type where the data — subject to the classification disclosure — suggests near-parity with existing member performance. PCL and ALR are the only other prospective products with sufficient volume to evaluate meaningfully; both convert at roughly 1/15th the existing rate, and the economics should be modeled explicitly before committing budget.
Conversion rate measures campaign activity in isolation. Market share — your closed loans divided by all loans closed in the same prescreened population — measures whether you are winning or losing relative to competitors. Only market share tells you if your competitive position is improving. Track it every campaign, every product, every segment.
Existing member campaigns outperform prospective campaigns by 12–25× across every well-evidenced product type — Personal Consolidation Loans, Auto Loan Refinance, Auto Loan Term Extensions, and HELOC Consolidation. Your members are in the market right now, and they are choosing other institutions when you don’t make an offer. Continuous existing-member coverage is the highest-return activity available to a credit union or community bank.
Prospective campaigns convert at 1/12th to 1/25th the rate of existing member campaigns — except for Mortgage New Purchase, where the gap narrows to approximately 1.3×. That exception is the acquisition entry point: win the mortgage, gain a member, then deploy the existing-member campaigns where your institution already dominates. Not every prospective offer type earns the economics; choose the ones where the data says you can compete.
Volume, conversion rate, and cost-per-acquisition are useful inputs. But the only question that answers “are we getting more competitive?” is: did our share of the prescreened market go up? Campaigns that increase market share are working. Campaigns that don’t — regardless of what conversion rate shows — are not.
Four products tell a consistent story: the existing member relationship produces 12–25x the market share of prospective campaigns. One product — mortgage new purchase — appears to be an exception, with a gap of approximately 1.3x, subject to pending classification review. That exception, if it holds, is the acquisition strategy: maximize existing member coverage on proven products, and use MNP prospective campaigns to bring in new members who will encounter those same high-converting campaigns for years — a flywheel that the Oliver Wyman research suggests is more accessible to credit unions, with their portfolio-and-service model, than to the large banks where the mortgage cross-sell thesis has historically underdelivered.
Micronotes helps credit unions and community banks run smarter prescreen campaigns — combining FCRA-compliant targeting, behavioral analytics, and campaign performance measurement to maximize market share across both existing and prospective segments.
Across virtually every prescreen campaign we analyze at Micronotes, the same pattern emerges: community financial institutions are losing to competitors in the higher income, higher credit score, and higher balance segments. These aren’t marginal prospects. They’re the borrowers every lender wants — and they’re slipping through CFI fingers at an alarming rate.
The irony is sharp. The consumers your institution is best positioned to serve — creditworthy, financially stable, community-rooted — are the same ones being aggressively courted by fintechs, mega banks, and marketplace lenders. And in many cases, they’re winning not because your rates are bad, but because the offer experience doesn’t match what a sophisticated borrower expects.
Let’s ground this in real numbers. We recently conducted a rate comparison for a credit union launching an auto loan refinance prescreen campaign. Their rates were strong — significantly below market averages reported by Experian’s Q4 2025 State of the Automotive Finance Market. For borrowers with 760+ credit scores, the credit union offered rates between 5.24% and 6.74% depending on term length, well under the Experian super-prime used car average of approximately 7.80%. For the 700–759 tier, their rates of 6.24%–6.99% sat meaningfully below the roughly 9.70% prime market average. At every credit tier, the institution was delivering real value.
But here’s where it gets tricky. When we benchmarked against the best advertised rates on LendingTree and Bankrate, the picture shifted. Top marketplace lenders were advertising floor rates starting around 4.65% for their most qualified borrowers — rates that, while not typical offers, are exactly what a rate-shopping super-prime consumer will find when they comparison-shop online. And that’s precisely what this segment does.
Higher-income, higher-credit consumers are financially sophisticated. They compare APRs. They use aggregators. They treat your prescreen offer not as a destination but as a data point in a broader shopping exercise. A Federal Reserve study on consumer financial behavior has consistently shown that higher-income households are more likely to shop multiple lenders before committing — and the proliferation of digital comparison tools has only accelerated this tendency.
Rate competitiveness is usually the primary driver, but it’s not the only one. Several interconnected factors explain why your highest-quality prospects disproportionately fund elsewhere.
First, offer amount mismatch. Higher-income consumers often have larger borrowing needs. If your prescreen offer is calibrated conservatively relative to what they actually need — or what a competitor is willing to extend — they’ll go elsewhere simply to get the loan size they want.
Second, speed and digital experience. This segment places a premium on frictionless, fast decisioning. If a fintech or large bank can approve and fund in 24–48 hours with a polished digital interface, that can override a modest rate advantage on your end. The traditional 21-day approval timeline that characterizes many credit union lending processes is a dealbreaker for borrowers accustomed to instant gratification.
Third, existing external relationships. Higher-income consumers are more likely to hold accounts at multiple institutions, including national banks with strong brand recognition. A prescreen from their primary checking bank often carries more psychological weight than one from a credit union they use less frequently.
Finally — and perhaps most importantly — the offer itself may not feel tailored. Higher-income consumers receive more financial solicitations than any other segment. If your prescreen letter reads as generic, it gets dismissed. Personalization signals — referencing an existing relationship, a specific loan purpose, a rate calculated against their actual credit profile — have a measurable impact on response rates. This is the difference between a piece of mail and a moment of recognition.
There’s a subtler dynamic at work as well: adverse selection in your response pool. Some of the higher-income consumers who respond to your offer and then fund elsewhere are doing so strategically — using your prescreen as a negotiating baseline, shopping your rate against competitors, and taking the better deal. This behavior is more common in higher-income tiers, where borrowers are savvier about leveraging competing offers. The result is that your campaign generates engagement but not conversion, inflating your marketing costs while the loan books elsewhere.
So what should community FIs do about it? We see two paths, and they lead to very different outcomes.
Path A: Differentiate the offer for this segment. Rather than competing purely on rate — a race that marketplace lenders with massive scale advantages will often win — layer additional incentives that create unique value. One of our clients is exploring a compelling approach: offering a sweepstakes entry or a percentage of the origination balance deposited directly into the borrower’s account as a closing bonus. This doesn’t require reducing the institution’s overall rate structure, which means existing pricing remains intact for the broader portfolio. But for the high-credit segment receiving the prescreen offer, there’s a tangible, differentiated benefit that an aggregator listing can’t replicate. Automated prescreen technology makes it possible to target these enhanced offers exclusively to the segment that needs them, avoiding the cost of blanket incentives.
Path B: Reduce volume to this segment. If higher-credit prospects are producing a disproportionate share of lost loans, the temptation is to simply mail fewer offers to them.
We strongly advocate for Path A. Retreating from your best prospects cedes the field to competitors permanently. Every high-credit borrower you choose not to pursue is one your competitor will gladly take — and once that relationship is established, winning it back becomes exponentially harder. The math favors finding a way to win, not finding a reason to stop trying.
This challenge is really a microcosm of the broader shift in prescreen marketing. The era of batch-and-blast — pull a list, mail an offer, measure response — is giving way to an era of precision-driven, segment-aware campaigns where different borrower profiles receive fundamentally different value propositions. Higher-income borrowers need to feel that the offer was built for them, not adapted from a template designed for someone else.
At Micronotes, our platform processes over 230 million credit records weekly, enabling the kind of granular segmentation that makes targeted incentive strategies operationally feasible. Post-campaign analytics can then measure exactly how different offer structures perform across credit tiers, creating a feedback loop that refines each subsequent campaign.
The community FIs that will thrive in this environment are the ones that stop treating prescreen as a single-strategy exercise and start treating it as a portfolio of segment-specific plays. Your highest-credit borrowers deserve your most creative offers — not your most generic ones.
The technology exists. The data supports it. The only remaining question is whether your institution will compete for its best prospects or concede them to someone else.
Community FIs chasing younger demographics may be ignoring their most profitable lending prospects: engaged 55+ members. Data shows this segment drives superior loan-to-share performance when properly cultivated.
Credit unions stress-testing their card programs focus on processors, margins, and scale—but overlook the critical question of acquisition infrastructure. Before choosing in-house or agent models, leaders must ask: where will your next qualified cardholders come from?
TD Bank’s billion-dollar AI strategy reveals a blueprint community FIs can follow: trust-first frameworks and human oversight turn AI-powered prescreen into a compliant growth engine. Here’s how to capture enterprise-level results without enterprise-level resources.
Prescreen marketing typically focuses on rates and credit limits. A shift toward outcome-based messaging—tying firm offers to member aspirations like homeownership or debt freedom—can unlock deeper engagement and differentiation.
Most banks have data and consumer permission to personalize, but remain stuck at broad segments. Learn why prescreen marketing offers community FIs a practical path to true 1:1 personalization.