The Tale of Two Diminishing Marginals

You're in your favourite café, first coffee of the day. The warmth spreads through your hands and the world sharpens. You order a second. It's good — not quite the first, but good. Somewhere around the third you notice you're jittery, vaguely regretful, and drinking mostly out of momentum.

That slide from bliss to regret has a name: diminishing marginal utility. And it has a twin that gets confused with it constantly — diminishing marginal returns. They describe the same shape of curve on opposite sides of a market, and telling them apart is the difference between "why won't customers buy more?" and "why isn't hiring helping?"

One curve, two stories

Both ideas say the same structural thing: the next unit adds less than the last one did. What differs is who the unit belongs to.

Switch between the two panels above and the shapes look almost identical. The difference is entirely in what you're counting.

Diminishing marginal utility: the buyer's side

Utility is the satisfaction you get from consuming something. The first cup is the peak. Each additional cup delivers a smaller increment of enjoyment than the one before it, until the increment goes to zero — and then negative, when you're so over-caffeinated the coffee is actively hurting.

This is why you stop buying. Not because you ran out of money, and not because the price changed, but because the value to you of the next unit dropped below what you're being asked to pay.

Two things follow from that, and they're the practically useful part:

Your personal demand curve slopes down because of this. When you see a demand curve in a textbook, this is the mechanism underneath it for any single buyer. You'd pay a lot for the first unit and progressively less for each one after — so the more units on offer, the lower the price has to be to move them all.

Quantity discounts are built on it. Buy-one-get-one-half-off works precisely because the second unit is worth less to you than the first. The discount is the seller meeting your declining willingness to pay rather than hoping you'll ignore it. Economists call this second-degree price discrimination: the price varies with the quantity you buy, and the buyer sorts themselves.

Diminishing marginal returns: the seller's side

Now run a bakery. One oven, one baker, 100 pastries an hour. Hire a second baker and output jumps. Hire a third and it rises again, though less. Hire a fourth and fifth and the kitchen gets crowded, people wait on the oven, and eventually another pair of hands makes things worse.

That's diminishing marginal returns: as you add more of one input while holding the others fixed, the extra output from each additional unit of that input falls.

The fixed part is the whole point. Returns diminish because the oven, the floor space, and the number of trays did not change when the headcount did. Add a second oven and the curve resets — you've changed what was fixed.

This is the difference between diminishing returns and economies of scale. Scale is what happens when you grow everything together and average cost falls. Diminishing returns is what happens when you grow one thing and the rest stays still. A business that confuses them tries to solve a capacity problem by hiring, and wonders why the payroll went up and the output didn't.

Why the confusion costs money

The two mistakes look like this in practice.

Treating a buyer problem as a seller problem. Sales flatten, so you push harder: more ad spend, more outreach, more discounting on the same product to the same people. But if your existing customers have simply reached the point where another unit isn't worth it to them, more pressure on the same segment won't move much. What moves it is a new segment, or a different product, or a reason for the next unit to be worth something again.

Treating a seller problem as a buyer problem. Output stalls, so you assume demand is soft and cut prices. But if the real constraint is that your one oven is running flat out, cutting price just means selling the same number of pastries for less.

The diagnostic question is short: what is fixed here, and for whom?

What to actually do with this

If you're buying, the useful move is to notice when you've crossed the line. The third coffee, the fourth streaming subscription, the tenth similar shirt — the spending feels the same each time, the satisfaction does not. You're not being frugal by stopping; you're being accurate.

If you're selling, there are two separate levers and they aren't interchangeable:

• To fight diminishing utility on the demand side, restore novelty or add value. Seasonal variants, new versions, tiered offers, bundles that combine things whose utility hasn't decayed. The point isn't the gimmick — it's that a genuinely different thing has its own first unit.
• To fight diminishing returns on the supply side, relieve the binding constraint. Work out which input is actually fixed, and invest there rather than piling more onto the input that's already abundant.

Both come back to the same discipline: knowing where "more" stops being better, and being willing to stop there.