Most deposit pricing is built around products.

We establish a rate sheet. We publish rates. We measure results. If balances grow, we assume the pricing was attractive. If balances leave, we assume the pricing was insufficient. The underlying idea is that changing the product rate changes customer behavior.

At a macro level it does. The problem is that customers are not all facing the same decision. In Price Optimization language, different decision space boundaries exhibit different response functions. Customer response tends to cluster around threshold values, and those thresholds change over time as rates, offers and market conditions change. Once decision boundaries are recognized, their response functions can be measured and modeled separately, improving the precision of both prediction and intervention.

Traditional pricing asks, “What rate should we offer?”

Decision-space optimization asks, “Which customers are facing decision spaces that are likely to produce outcomes we would like to influence?”

The difference is significant.

Suppose ALCO would like to extend term commitments. A traditional response might be to increase longer-term CD rates for everyone. That approach certainly creates incentives, but it creates them for customers who were already going to extend, customers who were never going to extend, and everyone in between. The institution pays for all of them.

A decision-space approach begins somewhere else. We identify customers whose current decision spaces already place them near the boundary between staying in their current term and moving to a longer one. These customers are already considering the outcome we prefer. They do not require a large intervention. They may require a very small one.

The management challenge becomes changing the value equation just enough to influence the decision.

What makes this particularly powerful is that rates are only one possible lever.

Customers do not only evaluate rates. Customers evaluate offers and costs.

An offer may contain a rate, but it may also contain liquidity features, service benefits, renewal privileges, relationship advantages, flexibility provisions, incentives or other elements that have value to the customer. If customers are making decisions against a set of criteria rather than a single rate, then we have more tools available than simply adjusting pricing across an entire portfolio.

This creates an entirely different view of optimization.

Instead of asking how to change a rate sheet, we begin asking how to selectively alter decision spaces for customers whose choices remain undecided.

Suppose the objective is to encourage migration from Money Market into CDs. We identify customers whose incremental benefit from moving is already close to a decision threshold. A small targeted incentive may move the decision.

Suppose the objective is to reduce migration into longer terms. We identify customers facing unusually strong incentives to move and selectively reduce the advantage associated with that alternative.

In both cases, we are not attempting to control behavior. We are selectively influencing the value equation at a decision boundary.

That distinction matters because broad pricing actions are expensive.

A rate increase applied across an entire portfolio affects every dollar receiving that rate. A targeted change applied only to customers whose decisions are still in play affects a much smaller population. The behavioral outcome may be similar. The cost often is not.

Viewed this way, price optimization shifts from portfolio management to decision-space management.

The institution still manages products. Treasury still manages funding. ALM still manages balance sheet risk.

But the optimization target changes.

We are no longer trying to make a product attractive to everyone. We are identifying customers approaching a decision boundary and selectively improving the attractiveness of desired outcomes. That is a fundamentally different way of thinking about pricing.

Instead of managing rates and hoping customers respond, we begin by understanding the decision space customers are facing. We measure the relative value of competing alternatives. We identify where decision boundaries exist. Then we use pricing and other available levers to selectively influence outcomes where influence is most likely to matter.

The customer still decides. But we are no longer managing products alone. We are optimizing the decision spaces that produce portfolio outcomes.