For community banks · Deposit retention

Your best depositor moves half her balance on a Tuesday. Your team finds out on Friday.

We build the defection-propensity model and the intervention workflow. Your deposit team sees the 15 or 20 relationships that need attention now. UDAAP-safe scripts. SR 11-7 governance for the scoring model. Your people own the system.

The depositor who leaves on a Tuesday

This is how top-decile depositors leave. Not with a phone call. Not with a complaint.

Your best depositor has been with the bank for 14 years. She holds a high-six-figure money market, two CDs, and an operating account for the small business. She has never complained. She has never asked for a rate match. She has never threatened to leave.

On a Tuesday, she moves half the money market balance to a competitor offering 35 basis points more. Your branch manager finds out on Friday — when the deposit report lands. By then, the balance is gone. There is no conversation to have. The relationship has already shrunk.

They rate-shop quietly, move quietly, and your deposit team sees it after the fact. Nearly nine in ten community bankers now cite cost of funds as a very or extremely important risk — up 41 percentage points since 2022. The pressure is real. But the problem is not the rate environment. The problem is that the deposit team has no systematic way to see defection before it happens. They are defending NIM with a rear-view mirror.

The long-standing customer who relocates half her balance costs more than the new customer you never acquired. She was already funded. She was already profitable. Replacing her balance means paying the prevailing rate to someone with no history at the bank. The cost-of-funds damage is not the rate you lost her at. It is the rate you pay to replace her.

What changes

Your deposit team identifies at-risk customers before they leave. Not after.

The system scores every top-decile depositor on defection propensity — how likely they are to move balances in the next 30 to 90 days. The scoring model runs on your DDA data. It watches for the behavioral patterns that precede departures: declining transaction frequency, balance step-downs, CD maturity approaching with no renewal conversation, rate inquiries at the contact center.

Your deposit team sees a ranked list each week. Not a thousand names. The 15 or 20 relationships that need attention now.

The retention approach is personalized. When your team reaches out, they do not improvise. The system produces a personalized retention approach for each customer — the right offer, the right channel, the right framing. Not a three-row save script from a 2018 training binder. A thousand save scripts against the behavioral economics of each customer. Your branch banker or contact center representative sees the specific retention path before they pick up the phone.

The team measures what happened. Did the customer retain the balance? Did the offer work? Did the relationship stabilize or continue to deteriorate? The deposit team tracks outcomes by customer segment, by intervention type, by branch. They adjust. They learn what works at your bank, not at a bank in a case study.

Your people run this system. The deposit team lead reviews the scores, assigns the outreach, tracks the outcomes. When the model needs recalibration — a new rate environment, a shift in competitive dynamics — your team knows how to update the parameters. The capability stays after we leave.

How the build works

Four to seven months. The scoring model and the operational layer.

The engagement runs four to seven months, depending on data readiness and the scope of the contact center integration.

01

Scope and data

Month 1

We work with your deposit team, your data team, and your core vendor to map the data environment. What does the DDA data look like? Where are the behavioral signals? What does the deposit team already know — and how are they acting on it today? This phase produces the model specification and the intervention framework.

02

Model build

Months 2–3

We build the defection-propensity model on your data. Not a vendor's pre-built model. Not a model trained on another bank's deposit base. A model built on your customers, your rate environment, your competitive market. Your deposit team participates in the build — they see the features, they understand why the model flags certain behaviors.

03

Scripts and intervention design

Months 3–4

Personalized retention approaches for branch and contact center, UDAAP-safe offer structures, escalation paths for high-value relationships. Your team reviews every script template. Your compliance officer reviews the offer language.

04

Pilot and measurement

Months 4–6

The system goes live on a defined segment. Your deposit team uses it. They make the calls. They run the playbook. We measure retention rates, cost-of-funds impact, and intervention effectiveness together. The team adjusts the approach based on what the data shows.

05

Handoff and training

Months 6–7

Your deposit team owns the system. They know how to read the scores, run the interventions, measure the outcomes, and recalibrate when conditions change. We document everything — the model, the methodology, the intervention framework, the measurement approach.

Governance is built in. The scoring model carries SR 11-7 documentation as standard — model inventory entry, validation framework, performance monitoring, and the examiner-ready narrative. Your CRO does not receive a separate governance project. The documentation is produced during the build, not after it. Retention scripts are reviewed against UDAAP requirements. The governance layer is part of the system, not a deliverable bolted on at the end.

What the team learns

This is not a system your team receives. It is a capability your team builds.

Your deposit team learns to read defection-propensity scores and translate them into prioritized outreach. They learn to evaluate intervention effectiveness and adjust the approach. They learn to identify when the model is drifting and how to request recalibration.

Your data team learns the model architecture well enough to maintain it. They understand the feature set, the scoring methodology, and the performance thresholds.

Your compliance team understands the SR 11-7 documentation and can update it as the model evolves. They know where the governance artifacts are because they helped produce them.

When we leave, the deposit team is not waiting for a consultant to tell them what to do. They are protecting deposits with a system they understand and control.

Protect your deposit base

Book a 90-minute working session.

Your top-decile depositors are the foundation of your funding structure. A system that identifies defection risk before the balance moves — and gives your team the tools to retain those relationships — is not a technology project. It is deposit defense.

You describe the deposit pressure. We describe how we would build the system for your bank. You leave with a written assessment within 48 hours.