Concept

The headline number

communication

Overview

A headline number is the single figure a whole analysis is compressed into for an audience — most often a levelized cost in $/t, a net cost after credits, a breakeven price, or a payback. It is the number a reader remembers and acts on, and presenting it honestly — with its band, its basis, and its drivers — is a communication act distinct from computing it.

Body

What it is. A model produces many numbers; communicating it forces a single one to the front to stand in for the whole analysis. The headline carries the conclusion — “this process makes ammonia at about X per tonne” — and everything else in the model exists to support, qualify, or defend it. Which figure becomes the headline (a gross cost, a net cost, a margin, a breakeven) is a presentation choice, not a fixed output.

It is inseparable from three things. A headline number alone is not interpretable; it travels honestly only carrying:

its band — the ±accuracy class of the estimate; a single point with no range implies a precision the analysis lacks (see false precision);
its basis — the system boundary, capacity factor, and annualization it was computed on, without which it cannot be compared to anything;
its drivers — the few inputs that move it, which say how robust it is.

A headline stripped of these is a number that cannot be defended, compared, or stress-tested.

Gross versus net is a headline choice. The same process has several legitimate headlines: a gross levelized cost, or a cost net of byproduct and policy credits. Which one is “the” headline is a presentation decision that can move the figure substantially, and the two are different numbers that must be labeled as such — a net headline beside a gross one compares unlike things.

One number hides structure. By construction the headline collapses the whole cost structure — the fixed/variable mix, the component shares — into a single figure. Two processes with the same headline can have opposite robustness: one fixed-cost-heavy and fragile to utilization, the other variable-cost-heavy and exposed to input prices. The headline is a pointer into the analysis, not a substitute for it.

Limits & typical error

A point headline implies a precision the estimate lacks. Quoting one figure with no band reads as exact; the accuracy-class band is part of the number, and dropping it is a false-precision claim by omission.
A headline without its basis is non-comparable. Two headline costs taken over different boundaries, capacity factors, or discount-rate assumptions are not the same measurement; unless the basis travels with the figure, “X beats Y” is unfounded — and the basis is the part most often left off.
Gross/net mislabeling overstates the case. Comparing a credit-netted headline to a gross one, or presenting a net figure without flagging the credits behind it, inflates apparent competitiveness and hides how much of the result depends on policy.
The single figure conceals the cost structure that sets its sensitivity. Because the headline hides the fixed/variable split, it cannot by itself say whether the result is fragile; two equal headlines can respond oppositely to the same price or utilization shock.
The freedom to choose a headline is the freedom to mislead with a true number. Selecting the metric, scenario, or basis that flatters — a best-case capacity factor, the net-of-credits figure, the one comparator the process beats — yields a defensible-looking number that misrepresents the central case without containing a single false digit.
A stated headline anchors the audience. Once announced, a headline number disproportionately shapes the reader’s takeaway; a rounded or early figure can dominate the conclusion out of proportion to the confidence behind it, which is why what gets chosen as the headline carries weight beyond its accuracy.

Mini-example

The honest headline for green ammonia is not “$805/t.” It is a sentence: ”~$800/t NH₃ (±~30%), gross of credits, on the power-to-ammonia boundary at capacity factor ~0.90, driven mainly by the power price.” That one line carries the figure, its band, its basis, and its top driver — everything needed to compare or challenge it. The same plant also has a legitimate net headline: after the ~$540/t clean-hydrogen credit, ~~$260/t — a different, equally valid number that must be labeled net and flagged as policy-dependent. Same plant, two headlines (~~$800 gross / ~$260 net), each true, each meaningless without its label.

Separately, to show a true number chosen to mislead: quoting the net ~$260/t alone, beside gray ammonia’s ~$200–400/t, to claim “cost-competitive today” hides that the gross ~$800/t is 2–3× the incumbent and that the entire gap is closed by a single time-limited credit. Every figure is real; the selection is the misrepresentation.