Case Studies / Community Bank
Banking & Financial Services ~$1.2B assets · 11 branches · Community bank

Local Community Bank

A community bank competing on deposit pricing across a regional branch network.

Lead service: Curated datasets + benchmarking + managed data + win-scenario modeling Engagement: 18 weeks

The situation

  • Deposit pricing was guesswork with no view of how it compared to peers.
  • Public data and proprietary research were never combined into one picture.
  • Board reporting was scattered and backward-looking.
  • No tool to test a pricing move before committing to it.

What Grounds did

  • Blended public APIs with curated proprietary datasets for deposit-share and pricing benchmarks.
  • Drilled benchmarks to the city and sub-market level.
  • Built rate-scenario and loan-pricing models with a win-scenario tool.
  • Ran it as a managed data service: Grounds-hosted, behind a read-only layer.

Capabilities deployed

Curated industry datasets Peer benchmarking Scenario modeling Win-scenario tool Managed data service

Results

~18 bps
Margin opportunity across the book
First
Peer-benchmarked deposit pricing view
1
Board synthesis layer
Hosted
Sticky, always-current data feed

What stays: A peer-benchmarked pricing view and a Grounds-hosted, always-current data feed.

Explore the model

Run your own numbers.

The same kind of live model we build inside the engagement. Move the inputs and watch the outcome change.

Interactive model

Why the data blend changes the answer

Public benchmarks set the peer rate; proprietary local data reveals how depositors actually behave when you price below it.

How to use: pick a segment, then set the rate environment, your offered rate, and balances. The curves are the annual value of pricing below peer, from public benchmarks alone vs blended with Grounds' proprietary local panel. They diverge because each assumes a very different deposit beta; the space between them is what the data unlocks.
4.10%
3.85%
$55M
National beta · public assumes
0.95
~24% runoff assumed
Local beta · Grounds measures
0.25
~6% actual runoff
In a 4.10% environment, holding 3.85% reads as -$156k on public data (it assumes ~24% of balances walk). The local panel shows retail core deposits are stickier, turning the same move into $60k.

Illustrative example, not a specific client engagement.

How the blend is built
Public APIs → peer rate & environment
▪Peer deposit rates: FDIC call reports & SOD
▪Rate environment: FRED / Treasury curve
▪Local market demographics: Census
Grounds proprietary panel → local beta
▪Branch-level local competitor rate sheets
▪Observed local deposit flows by vintage
▪Account-level switching behavior
Public data can't see local switching behavior, so on its own it falls back to the national beta, and overstates how many depositors would leave.
Grounds · scenario printout
SegmentRetail core deposits
Offered rate3.85%
Peer median (public)4.10%
Local fair rate (proprietary)3.80%
Deposit retention94%
Value captured (yr 1)$60k
Unlocked by the data blend$217k
the decision swing only visible with proprietary local data
Grounds engagement (illustrative)$7k
~12% of modeled value · priced on outcome, not hours
Applied across the bank's ~$1.2B book, disciplined sub-market pricing scaled to ~18 bps, about $2.2M/yr.

Recognize your business here?

Tell us your industry and the decision you're sitting on. We'll show you the version of this model built around your numbers.

Start with your industry →