From Quarterly Rate Sheets to Customer-Level Decisions

Deposit pricing used to be a quiet, predictable exercise. A treasury team reviewed competitor rates, adjusted a handful of posted tiers once a quarter, and relied on customer inertia to do most of the retention work. That model is being dismantled in front of us, and the banks moving fastest are not just tweaking it but replacing it entirely.

Simon-Kucher’s 2026 deposit management analysis captures the shift bluntly. The most advanced institutions are using AI and behavioral analytics to understand how money moves between accounts, which channels customers use, and what patterns precede attrition or growth. That capability lets them act before balances walk out the door, so rather than reacting after outflows, these banks predict which customers are likely to move funds and proactively adjust rates or offers.

The practical effect is that deposit pricing is no longer a product decision made at the balance sheet level, but a customer decision made continuously at the account level. For retail banking executives, that is both opportunity and risk, because the same precision that protects margins can also strain the primary relationships that hold a deposit base together.

Why This Is Happening Now, After Years of Flat Pricing

Three forces have converged to push banks toward dynamic deposit pricing in 2026, and none of them are reversing. The first is margin compression. With the Federal Reserve easing and deposit betas catching up with lending betas, McKinsey has projected retail banking margin declines of 5% to 10% across geographies through the year, which means every basis point on the deposit side now matters more than it did two years ago.

The second is the data foundation. Most large U.S. banks now have enough clean transaction history, channel telemetry, and rate-change records to run the kind of elasticity models that used to live only in treasury. Simon-Kucher notes that roughly 24 months or more of transaction histories, rate changes, and competitor benchmarks are typically the minimum needed before AI can personalize pricing decisions, and many banks have finally crossed that threshold.

The third is competitive pressure from the agent and aggregator economy. Consumer AI agents and comparison tools can surface better rates almost instantly, which means banks paying a flat rate to millions of customers are simultaneously overpaying the insensitive ones and underpaying the ones about to leave. Dynamic pricing is the rational answer to that misallocation, and it is why Accenture and Deloitte have both flagged it as a 2026 capability, not a future one.

What the Upside Actually Looks Like

The economics are hard to argue with when the models are set up correctly. Simon-Kucher’s case work suggests pricing optimization alone can deliver 8 to 18 basis points of margin uplift, often without other measures, and the firm has pointed to one large North American bank that generated more than $45 million in deposit cost savings while keeping portfolio volumes unchanged. Those numbers tend to catch the attention of chief financial officers who have watched flat pricing quietly leak margin for years.

The mechanics vary, but the pattern is consistent. Behavioral segmentation identifies customers whose balances are price-sensitive and those whose balances are not. Rate offers are targeted at the sensitive segments, often with time-bound conditions like hold a balance steady for 90 days and earn a higher rate, while insensitive customers continue on posted rates. Campaign spend shifts from blanket promotions to precision nudges, and the operational work moves from manual discount approvals toward digitized workflows with clear rules for when exceptions can be granted.

Executives who have run this play describe the real win as not maximum yield, but better yield for the same cost. Reallocating rate incentives toward the customers who actually respond to them lets banks hold volume, protect funding stability, and reduce the blanket cost of deposit growth in one move.

Where Dynamic Pricing Starts to Hurt the Relationship

The risks are not the ones finance teams usually think about first. The obvious pitfalls, such as model error, data quality, and rate-cannibalization across products, are real but solvable. The quieter risks sit on the customer side of the house, and they deserve board-level attention.

The first is household consistency. If a bank runs customer-level pricing without a view of the full household, two spouses with shared finances can end up on different rates for the same product, which is exactly the kind of surprise that turns a primary relationship into a shopping exercise. Simon-Kucher specifically flags this, arguing that fairness and consistency must be maintained across customer relationships so that households, client constellations, or corporate affiliates receive coherent treatment.

The second is disclosure. Customers have generally accepted that credit card interest rates vary by risk profile, but the expectation on the deposit side has been closer to flat. A customer who learns through a friend or a comparison site that the bank is paying a neighbor a higher rate on the same product will often respond badly, even if the model has a perfectly defensible reason for the difference. Banks that handle this well are investing in clear customer-facing language about how rates are set and are building in-app visibility into current and potential offers, which tends to reduce the sense of arbitrariness.

The third is fair-treatment scrutiny. Regulators have not issued specific deposit pricing guidance that mirrors fair-lending rules, but the direction of travel is visible. Institutions treating dynamic pricing as a governed product with documented inputs, explainable outputs, and disparate-impact testing are in a far better position than those running it as a black-box optimization. The same controls that make the model defensible also make it better, because they force the team to understand which variables are actually driving decisions.

What It Takes Operationally to Do This Well

Dynamic deposit pricing is less a technology purchase than an operating-model change, and most banks that struggle with it are tripped up by the operating model, not the math. The capability stack typically runs across four layers.

Data is the foundation. Before any AI can personalize pricing decisions, the bank needs clean and consistent transaction data, rate history, and competitor benchmarks at scale. Simon-Kucher’s recommendation of 24 months of data as a floor is a useful benchmark. Decisioning sits on top of that foundation and includes the elasticity models, attrition-prediction models, and portfolio-level constraints that turn customer signals into specific rate recommendations.

Delivery is where many programs fall apart. A recommendation that cannot reach the relationship manager, the call-center agent, or the digital channel in real time is worth very little. The banks that execute well have plumbed pricing recommendations directly into the front-line tools their people actually use and have standardized the discount-approval workflow so that exceptions are tracked, not buried in email threads.

Governance closes the loop. That means a pricing committee with real authority over when personalization is used, clear rules for household and relationship consistency, documentation that can survive a regulator’s read, and monitoring that flags drift in the models. Banks that skip this layer tend to produce strong short-term margin numbers, followed by awkward retroactive conversations.

How to Keep the Primary Relationship at the Center

The deepest risk in dynamic pricing is that the bank optimizes for individual rate sensitivity and slowly trains its customers to behave more transactionally. McKinsey’s long-running work on primary relationships has been consistent on this point, arguing that customers with genuine primary relationships keep most of their deposits in place and remain open to additional products and services. A pricing engine that treats those customers purely as rate optimization problems can quietly unwind that dynamic.

The institutions handling this best treat dynamic pricing as one lever inside a broader relationship strategy, not the strategy itself. That shows up in practical choices, including rate offers bundled with advice or service upgrades rather than delivered as standalone coupons, household-level views that prevent inconsistent treatment across related accounts, and explicit segmentation that protects long-standing customers from feeling pushed into a continuous negotiation.

The near-term agenda for retail banking leaders is concrete. Build the data foundation if it is not already there, stand up a pricing committee with clear governance authority, pilot personalization inside one product and one segment before scaling, and measure outcomes not just in margin basis points but in primary-relationship retention and cross-hold rates. Done right, dynamic pricing protects the deposit franchise in a tightening market. Done without those guardrails, it can unwind years of relationship-building one finely tuned offer at a time.