Utility Maximization

Utility Maximization is what happens when your Utility Function meets your Budget. You already know your Utility Function ranks outcomes by what you care about. You already know your Budget caps how much you can spend. Utility Maximization is the process of finding the specific Allocation - the exact dollar split across categories - that produces the highest possible utility score without exceeding your Budget.

The core mechanic: keep shifting dollars from lower-value uses to higher-value uses until no shift can improve the total. When investments can be sized at any dollar amount, the optimum has a clean property - every funded category ends up with the same marginal value per dollar. If Marketing Spend's next dollar would generate more utility than tooling's next dollar, you haven't maximized yet. Move money toward marketing.

Most real investments come in fixed increments. You hire a person or you don't. You fund a campaign at its minimum effective level or skip it. With discrete investments, you rank all available increments by marginal value per dollar and fund them in order until the Budget runs out. This ranked approach works well when each increment is small relative to the Budget - the approximation error is minor. But when increments are large (a single hire consuming 40% of Budget), the ranked order can miss better combinations. Funding a high-ranked but expensive item early can strand enough Budget to waste more value than a different combination would. When the ranked order strands a large fraction of Budget, test alternatives: check whether swapping the last funded items for the best unfunded one improves total utility.

This is not the same as picking the single best project. It's about how much goes to each competing use when you're funding multiple things simultaneously.

Every P&L owner faces the same structural problem: more good ideas than money. Utility Maximization is the framework that turns 'I have five good projects and Budget for three' into a principled Allocation instead of a political one.

Without it, you get two failure modes:

1. 1)Peanut-butter spreading - giving every initiative a little funding so nobody complains. This ignores diminishing returns and wastes Budget on low-marginal-value uses.
2. 2)All-in bets - dumping the entire Budget into the highest-ROI project. This ignores the fact that returns diminish as you scale any single initiative. The 10th sales hire doesn't produce the same Revenue as the 1st.

Operators who internalize Utility Maximization make better resource allocation decisions because they're reasoning about where the next dollar creates the most value, not where the average dollar looks best.

The mechanics rely on one key insight: diminishing returns. Almost every business investment produces less incremental value as you pour more money into it. Your first $50K in Marketing Spend might generate $200K in Revenue. The next $50K might generate $120K. The next, $60K.

Here's the process:

1. 1)List your competing uses. These are the categories your Budget can fund: Hiring Targets, Marketing Spend, tooling, capacity expansion, and so on.

2. 2)Estimate the marginal value curve for each use. For each category, ask: if I spend another increment here, how much does my utility score improve? What about the next increment? The answers should decline - that's diminishing returns.

These estimates are uncertain, but you're not guessing blind. For sales hires, look at your last few hires' ramp-to-Revenue curves and use the most recent hire's contribution as your prior for the next one - if hires #3 and #4 ramped to $250K and $200K in first-year Revenue respectively, $150K is a reasonable estimate for hire #5. For Marketing Spend, measure your Marketing Spend per Pipeline Volume lead at your current spend level, then check whether doubling spend in a channel last time produced anywhere near double the leads (it usually didn't - that's the diminishing curve). For newer investments without history, use last quarter's actual marginal returns from the closest comparable category as a starting prior, and flag the estimate as low-confidence.

When two options are close in estimated marginal value per dollar, run a Sensitivity Analysis: check whether your Allocation changes if the estimates shift by 20-30%. If it doesn't, the Allocation is robust. If it does, gather better data before committing.

3. 3)Normalize before combining. If your Utility Function scores multiple objectives measured in different units - Revenue in dollars, Pipeline Volume in lead counts, Churn Rate in percentages - put them on a common scale before applying weights. Without normalization, the dimension with the largest raw numbers drowns out everything else, and your weights become decorative. Common approaches: scale each dimension to a [0, 1] range (divide each value by the maximum value across all increments), or convert all dimensions to dollar equivalents (e.g., each Pipeline Volume lead is worth approximately $X in long-run Revenue conversion).

4. 4)Rank by marginal value per dollar, then allocate. Fund the highest-marginal-value increment first. Continue down the list until your Budget is exhausted or no remaining increment fits within the remaining Budget.

5. 5)Check the result. This is the step most people skip. If the unspent Budget is small relative to the cheapest unfunded increment - say, $5K stranded against a $25K minimum - the waste is minor and the ranked order is likely optimal. But if the stranded amount is a large fraction of Budget, test alternatives: try replacing the last one or two funded increments with the best unfunded one and compare total utility. The worked examples below show both cases - one where the swap test finds a better Allocation, and one where it confirms the ranked order was right.

Interaction effects. The process above treats each category's marginal value as independent of spending in other categories. In practice, categories can be complements. Marketing generates Pipeline Volume that sales hires convert. A second sales hire's marginal value depends on whether Marketing Spend is generating enough Pipeline Volume to keep both hires productive. When two categories are strongly complementary, estimate their joint marginal curve - 'what does this sales hire produce at our planned level of Marketing Spend?' - rather than treating each category in isolation. The independent model works well when categories are loosely coupled. When they interact strongly, the joint estimate prevents misallocation.

Shadow Price. The Shadow Price of your Budget is the value of relaxing the constraint - how much more utility you'd get if your Budget grew. When investments can be sized at any dollar amount, this equals the marginal value of one more dollar, and it's always well-defined. With discrete investments, the Shadow Price is a step function: if the cheapest unfunded increment costs $80K and you have $45K stranded, one more dollar doesn't help. Neither does $34K more. The Shadow Price is effectively zero until you accumulate enough additional Budget to fund the next increment - in this case, $35K more. At exactly that threshold, it jumps to the marginal value per dollar of the newly fundable increment. This matters when deciding whether to fight for incremental Budget: a small increase may buy nothing, while a specific threshold amount unlocks a discrete jump in utility.

Use Utility Maximization whenever you face a constrained Allocation decision - multiple uses competing for a single pool of money, time, or capacity.

Good fits:

• •Quarterly Budget Allocation across departments or initiatives
• •Splitting Marketing Spend across channels with different diminishing returns curves
• •resource allocation when you have more projects than engineers
• •Capital Allocation decisions across Capital Investment options
• •Deciding how much of your Budget to reserve vs. deploy (the reserve is itself a use with its own marginal value - optionality to fund unexpected opportunities has real value via risk appetite)

Poor fits:

• •Binary go/no-go decisions with no partial funding option (use Expected Value instead)
• •Decisions where your Utility Function is just 'maximize dollars' with no other objectives (use ROI ranking instead)
• •Situations where you genuinely can't estimate diminishing returns for the competing uses

The framework is most powerful when your Utility Function weights multiple objectives - Revenue growth and risk appetite and Time Horizon preferences - because that's when pure ROI ranking breaks down.

Utility Maximization is the operational payoff of the two concepts you've already learned. Your Utility Function defines what you're optimizing - the Scoring Model. Your Budget defines the constraint - how much you have to work with. Utility Maximization is the process of solving for the best Allocation given both. Downstream, this connects directly to Shadow Price (the value of relaxing your Budget constraint - but remember it's a step function with discrete investments), marginal dollar allocation (deciding where the next dollar goes), and Zero-Based Budgeting (rebuilding the optimal Allocation from scratch each period instead of rolling forward last period's split). It also underpins Capital Allocation at higher stakes - the same ranked-marginal logic with swap-test discipline applies whether you're splitting a $200K quarterly Budget or a $50M Capital Investment pool across PE portfolio companies.