Deaths Are at a Record High. Funeral Homes Are Closing.

In 2025, 363,400 people died in Korea. Outside the pandemic years, that is the highest figure on record. Aging demographics guarantee the number keeps climbing. Demand this legible almost never exists.

So the funeral business should be booming. It isn't. Funeral homes nationwide fell from 1,107 in 2021 to 1,075 in 2025. Demand is rising on a near-certain trajectory, and operators are walking away from it.

Strange, but not rare. Nursing hospitals filling with an aging population, farms in a bumper-harvest year, subway systems carrying more riders — all standing in the same place. So is the market-size slide we look at several times a day.

Rising Demand, Shrinking Supply: Who Actually Captures the Volume

In May 2024, we published a piece on the time-lagged doom loop between Korea's private education market and its collapsing birthrate. The question was why the tutoring market keeps growing while the number of children shrinks. Students — the Q — are clearly declining, but spending per child is climbing faster, so the market expands anyway. Read the single macro headline and conclude "education is finished," and you land on precisely the wrong answer.

This piece is the mirror image. If Q can fall while a market grows, Q can rise while a market doesn't. The second case is the more dangerous of the two, because being right about demand is exactly what keeps drawing new entrants in.

Business profit reduces to a simple expression:

Profit = (P − C) × Q

What we watch hardest is almost always Q. Users, transactions, footfall — numbers that are visible and that move when you push them. Market sizing runs on the same instinct. "The elderly population is growing." "Single-person households are growing." Every one of those statements is a statement about Q.

Look at the equation again. Q sets the magnitude of profit. It cannot set the sign. The sign belongs to the bracket, P − C. If margin is negative, a larger Q produces a larger loss. Scaling a negative-margin business is not acceleration; it's descent. And converting a larger Q into a higher price requires one condition: the buyer has to be the party that needs the deal more.

This piece is about the opposite condition — businesses where sellers are scarce and pricing power still isn't theirs. There are three types. In the first, you hold the price list and it still doesn't pay. In the other two, you never held it at all.

Type 1: Pricing Power Is Worthless When Revenue per Transaction Collapses

The first type never requires changing the price list. Prices hold. Customers simply stop buying.

Where funeral home revenue actually came from

Funeral home revenue comes from three lines: room rental, food and beverage, and supplies such as caskets and burial garments. Historically the largest of the three was hospitality — F&B. Room rental carries a fixed cost basis and supplies have predictable unit economics, but when mourners arrived in volume and stayed long, F&B revenue compounded on top.

That structure has broken. No-visitation funerals and family-only services have risen sharply, and by industry accounts the number of mourners per funeral home has fallen to less than half its level a decade ago. Wiring a condolence gift instead of attending has become routine. Declining alcohol consumption flows straight through to the same line. Funeral service providers have halved as well — from 163 in 2017 to 76 as of March 2026.

Note what P means here. It is not the sticker price of a room or a plate of food. If Q is defined as one funeral, then P is the average revenue realized per funeral.

Counting cases and counting revenue per case are different exercises. Deaths are rising, so funeral volume is expected to rise with them. But if realized revenue per funeral falls faster, the economics of the market deteriorate regardless.

Fixed costs don't shrink with the guest list

The count goes up. What used to be sold inside each count disappears.

This is what makes the type difficult. Nobody cut a price. The same goods at the same prices are simply purchased less, and the traditional funeral itself has shrunk in scale. Meanwhile, cost does not follow revenue down. Facilities and staffing do not contract in proportion to a smaller guest list. Recoverable revenue per funeral falls, cost holds, and P − C deteriorates. Unless the expected increase in Q offsets that gap, profitability falls with it.

And because the problem was never a missing pricing right, raising prices does not solve it. Raise room rental and you get more no-visitation funerals. The price list nominally belongs to the supplier. Which line of it the customer chooses to read does not.

Type 2: When a Single Payer Writes Your Price List

In the second type, pricing authority sat outside the business from day one. The buyer writes the list, not the seller — and there is effectively only one buyer.

Nursing hospitals: a buyer you cannot refuse

Nursing hospital reimbursement has been set on a per-diem flat rate since 2008. Under a per-diem system, the same patient grade pays the same amount no matter how much work the patient requires. Deploying more medication and more staff to a higher-acuity case earns no corresponding fee. The path of growing revenue by increasing input is closed by design.

The nursing hospital's customer is not the patient. It is the National Health Insurance Service, Korea's single public payer. And because medical law prohibits refusing treatment without justifiable cause, loss-making customers cannot be screened out either. Between 2021 and 2025, 273 nursing hospitals opened and 441 closed — inside a market whose demand curve is underwritten by demographics.

Pediatrics: the supply side broke first

Pediatrics faces the same buyer and reaches the same conclusion from the opposite direction. The pediatric population has shrunk, yet parents line up at opening hour for an appointment. Not because demand exploded, but because supply collapsed first. Reimbursement rates sit at the bottom of every medical specialty, and once vaccinations were absorbed into the national immunization program, the revenue base narrowed to consultation fees alone. In 2024, 59 clinics opened and 89 closed — the highest closure rate of any specialty.

Utilities and transit: same structure, larger balance sheet

Electricity and transit stand in the same position. When input costs spiked in 2022, KEPCO posted an operating loss in the KRW 32 trillion range; the moment it was permitted to raise tariffs, it swung straight back to profit. Seoul's subway loses KRW 781 on every passenger it carries.

Type 3: When Market Structure Sets the Price and No One Signs It

In the third type, there is no actor pressing prices down. The structure of the market does the work.

The paradox of the bumper harvest

Farming is the canonical case. When a good harvest reduces farm income, we call it the paradox of the bumper harvest. Produce is a staple, so people don't eat meaningfully more of it because it got cheaper. A small increase in supply is enough to send farmgate prices down hard. Yield rises 10%, price falls by half, and income drops. Growing Q was the mistake.

Why the shelf price holds when the farmgate price halves

The obvious follow-up: if farmgate prices fall that far, why doesn't the grocery shelf? Because half of what the consumer pays was never the farmer's to begin with.

As of 2024, the distribution cost ratio for agricultural products is 49.2%. Pay KRW 10,000 at retail and KRW 5,080 reaches the farm. Nor is the remaining KRW 4,920 anyone's clean margin. KRW 1,430 is profit across the distribution chain. KRW 3,490 is cost — KRW 1,650 in direct costs like harvesting, packaging, and transport, and KRW 1,840 in indirect costs like labor and store upkeep. By stage, the weight concentrates at the end of the chain: 9.4% at shipping, 14.2% at wholesale, 25.6% at retail.

The critical point is that these costs move almost independently of the farmgate price. Cabbage halving in value does not halve logistics or wages. Korea's National Assembly Budget Office has noted that these direct costs are strongly fixed in character, which limits how far they can be compressed. Distribution players such as Nonghyup, the national agricultural cooperative, further smooth the amplitude in both directions. What ends up falling hard is only the price the farmer receives.

Why Startup TAM Slides Are Really Just Q

Run diligence long enough and you see market-size slides several times a day. The number on them is usually built the same way: count the eligible population or transaction volume, multiply by today's average price. The calculation treats price as a constant. Q is sent into the future; P is left behind in the present.

"Aging drives deaths up, so the funeral market grows" is precisely that calculation. And as we've seen, it's wrong.

Which is why the questions we put to founders tend to come after Q, not before it.

When the answer isn't P, reshape C

Finally, the answer doesn't always live in P. Dawn delivery was a business where variable costs grew faster than volume did. Kurly ran at a loss for most of its existence before posting its first half-year profit in the first half of 2025. A significant driver: fulfillment centers and cold-chain infrastructure moving into their depreciation phase, allowing the same asset base to absorb far more volume. If you can't raise P, reshaping C is still open. With one condition attached — you need enough capital to survive until you get there.

About Kakao Ventures

Founded in 2012 and backed by Kakao — Korea's leading tech platform — Kakao Ventures is one of Korea's most active Seed-stage venture capital firms, with approximately $280M USD in AUM. We partner with founders before the path is fully defined, when conviction in people matters more than proof in numbers.

Our portfolio includes Lunit (AI cancer diagnostics), Rebellions (AI semiconductors), and Dunamu (operator of Upbit, one of Asia's largest crypto exchanges).

If you're building at the edge of what's possible — we'd like to hear from you.