Death-Care Demand Forecast

How many people will die in a county each year, from now to 2040 — for every one of the 3,144 US counties. Delivered into Claude and other AI assistants over MCP, or called directly as an API.

Try it before you pay

No sign-up, no assistant needed: we publish free, complete pages for the 25 largest US counties, each showing expected annual deaths now and projected to 2040, with the method and its limits stated. See the free county pages — for example Los Angeles County, Cook County or Harris County.

If you prefer to query it directly: connect an MCP-capable assistant to product.kcalbinmcp.uk/mcp and query death-care demand immediately. No account needed. County names stay masked until you subscribe — you can confirm the data exists, and see the population and rate, before paying for the detail.

Plans

Six ways to buy, because the buyers are genuinely different.

County Report — $249, one-time

For an independent funeral home or cemetery asking a one-time question: should I expand here? One county — current-year expected deaths, the trajectory to 2040, and where that county ranks nationally for growth. Not a subscription; you ask this once.

Buy the County Report — $249

Professional — $2,250/year

• Every county, not just one
• Year-by-year projections 2025-2040
• National growth ranking against all 3,144 counties
• Multi-county and whole-state comparison

Subscribe — $2,250/yr

Supplier — $5,000/year

For casket, monument, urn and crematory-equipment makers and distributors. Professional access shaped for territory design and sales-force allocation across the 15,401 US funeral homes.

Subscribe — $5,000/yr

Institutional — $17,500/year

For investors, private equity and analysts. Everything in Professional plus bulk extract for your own modelling, and terms permitting extracts in investment-committee and LP materials with attribution.

Subscribe — $17,500/yr

Enterprise — $27,500/year

For national operators and consolidators. Bulk extract of every county across every year, internal redistribution across operating subsidiaries, higher rate limits.

Subscribe — $27,500/yr

Academic & Media — free

Free for academic research and editorial use, with attribution and no redistribution. If you are writing about how mortality is changing in a particular county, ask and we will send you the numbers.

How it works

Subscribe on a plan above and receive your API key by email. Add the server to your assistant — in Claude Desktop: Settings → MCP Servers → Add Server → URL https://product.kcalbinmcp.uk/mcp, then your API key. Query death-care demand directly in your conversations, or call the API from your own agent with a Bearer token.

What this forecast is, and is not

What you are buying is the projection layer. The underlying Census population projections and age-specific mortality rates are free public data. The value is the county-level allocation, the year-by-year trajectory to 2040, and the national ranking — assembled, calibrated and validated.

The projection is backtested. Tested against actual recorded deaths, averaged over three tests. In each test the county's share came only from that year's actual deaths. Run the way the product actually runs — without knowing the national total for the year — the typical county came out 7.82% from the actual figure one year ahead. The national path and the ageing pattern come from an earlier model built on data covering 2020 to 2024, which overlaps the test years, so these are not tests against a model blind to them.

How far ahead this figure actually is: The figure shown for this year is built from the most recent year of actual recorded deaths, which is two to two and a half years earlier. Tested at that distance the typical county came out 14.54% from the actual figure.

Which way it tends to be wrong: In all 5 tests the national projection came out higher than the actual number of deaths — by +1.1% to +15.6%. Expect this model to lean high.

How far the evidence reaches: These error figures cover one and two years ahead only. Nothing beyond two years has been tested against actual deaths, so the path out to 2040 is less certain than these numbers suggest.

If the national total were known: If the national total for the year were known, the typical county would be 5.03% out and 77% of counties within 10%, both averaged over the same three tests, with the typical state 1.13% out in the most recent test. The model this replaced was 19.18% typical with 26% within 10%, on the same three-test basis.

The caveat that matters: How many people die nationally is the largest source of error here. In the Census year 2022–2023 (Census years run July to June) the national projection came out +11.4% against actual, and the typical county error rose to 12.49%.

Error grows with distance. Error grows with distance. Against California's own free county projection this model sits within about 3% through 2035 and about 5.0% below it by 2040. That California figure is a comparison with another projection, not a measurement against actual deaths. It does not claim any single county is precise.

Large counties project more accurately than small ones, though less so than it first appears: run without knowing the national total, counties with over 5,000 deaths a year came out 4.55% from actual and those under 20 deaths 23.79% — though that is about 3 deaths.

Calibration is disclosed. Before anchoring, the model implied 3,142,860 US deaths for the Census year it is calibrated on; the actual recorded figure was 3,086,925. Every county is therefore scaled down by a factor of 0.98220, and that factor is stored on every row of the table these pages are built from.

A demand projection is a planning aid, not a guarantee about any specific county or year. It projects deaths, not market share, revenue, or the number of families who will choose any particular provider.

Built for death-care operators (market sizing and expansion planning) and for investors and analysts (due diligence and market sizing on death-care-adjacent assets).

Delivery and refunds

Digital delivery over MCP or the API. There is nothing to ship and no shipping cost. Subscription plans carry a full refund within 14 days of the first charge; after that you may cancel at any time, and cancellation stops future billing but is not retroactive. The one-time County Report is excluded from the refund window — it is a digital report delivered in full at purchase.

Need something else?

Custom geographic coverage, procurement paperwork, or a question before you subscribe.