Giovannoni: On the discipline public radio lost — and the metrics needed to recover it

Consider the listener-hour: one person listening for one hour, 60 minutes of human attention freely given. It is a real number, drawn from Nielsen’s measurement of broadcast and select streams — not a projection modeled from download counts, not a proxy based on signal strength or coverage.

We invented the listener-hour when we needed a sound basis for measurement, analysis and decision-making. We minted it as the currency of public radio’s programming — a coin whose two faces denominate the same hour of attention: as the service the programming gives the listener, and as the support the listener gives back. This is the virtuous cycle that powers the business of public radio:

Public service begets public support, which in turn begets public service.

The hour of human attention is one of the most useful measures we have of our significance. Millions of options now compete for it. When people choose public radio, they select it over every other claim on their time. That choice is a measure of the program’s significance. The amount of attention significant programming earns — counted in listener-hours — is how we assess our public service. And by proven extension, it is the reason listeners give us their money as freely as their time.

Taking the measure

Eric Nuzum has shown how the listener-hour can normalize attention to what a station offers via broadcast and all other platforms, the goal being to provide comparable measures of the attention each earns. At one station that prided itself on digital leadership, broadcast generated 99.94% of all audience attention, and everything digital combined for 0.06% percent. The highest digital share he identified among other public stations was about 2%; that’s 98 hours of attention generated by broadcast for every two hours generated by everything else the station produces.

Nuzum has not published the cost of serving one listener for one hour on each platform. But it is a number every public station should calculate. It is particularly useful in calculating the net return per listener-hour, by subtracting the cost per listener-hour from the financial return per listener-hour. A positive number indicates the programming is sustainable and generating surplus for other activities. A negative number indicates the endeavor is not financially sustainable as currently configured and is draining funds from other activities.

Public radio’s essential challenge is to sustain itself by providing the best public service it can afford. Yes, underwriting dollars have migrated to digital media — but they have migrated faster than the attention they presume to buy has. And yes, listening to terrestrial broadcast is in decline, but reports of its death are greatly exaggerated.

FM still reigns as the center of listener attention in today’s hyper-fragmented audio market. Measuring the support that attention earns, though, is far more complicated than measuring the attention itself.

It’s trivial to calculate the cost of a radio program per listener-hour, and just as easy to calculate underwriting’s return. What’s difficult and expensive is linking listener support to listening at the program level. Seemingly obvious solutions fail. Dividing the pledges received during a program by its listener-hours doesn’t come close. Dividing a station’s total listener support by its total listener-hours does yield an average return per listener-hour across the schedule — a useful figure in itself — but some programs matter more to listeners than others, and that assessment of significance translates directly into giving.

Listener support of a station is a function of listeners’ use of the station in its entirety. That support must be allocated across the programming they hear. AUDIENCE 98 was the last study to explicitly link the value listeners placed on specific programs to the support their listening returned. It quantified the direct returns yielded by programming investments in both public service and financial terms, at both the national and local levels.

In 1999, NPR incorporated that knowledge into a powerful new pricing model that, for the first time, aligned the interests of the network, its stations and their listeners. The new “value-based” model charged a station not for what it could afford, as all previous formulas had done, but for what its programming was worth to listeners. AUDIENCE 98 had traced the source of that worth to personal importance — a listener’s belief that the program “is important in my life. I would miss it if it went away.” That conviction predicted giving to the station by listeners, which, in combination with the underwriting generated during the program, established the “listener-sensitive” value of the program to the station.

AUDIENCE 98’s measure of each program’s listener-sensitive value per listener-hour set its price. Each station paid NPR a fraction of the support its listening earned — a penny or two against several cents harvested. The fee was generous: The station kept the larger share. It was sustainable: NPR never charged more than the programming generated for the station. And it energized the virtuous cycle — the more significant programming a station carried, the more attention listeners gave to it (in listener-hours), the more the station earned to subsidize its own programming, and the more NPR earned to improve, or to make more of, its programming.

With the virtuous cycle of public service, public support baked into the model, programming became an investment with real, quantified returns — not just another expense. For nearly 20 years, in one form or another, the model served public radio and its listeners well … until it was abandoned.

Losing the measure

The model was not refuted; it was forgotten. Two lapses explain the decline.

The “public service, public support” economy is not taught in school. It must be explained to each new cohort of managers, especially those arriving from commercial enterprises. Thirty years after its invention, those who created it have retired or moved on, as have the first generations of managers taught to use it. The doctrine lost its continuity. Analytical services that once put the metrics of programming economics on scores of managers’ desks are today virtually unused.

The tools did not stop working, and the constructs did not stop being true. What faded was the common understanding of why they mattered.

Worse yet, one key metric went dark. AUDIENCE 98 quantified each program’s returns per listener-hour in the late 1990s. But it was never refreshed. No comparable study has re-established what an hour of listener attention is worth. A return unmeasured is a term missing from our calculus. As the number faded, stations’ perceptions of NPR’s fees devolved towards their pre-1999 status: expenses with no direct returns.

NPR’s subsequent models are best understood not as repudiations of the method but as consequences of that missing measure.

NPR’s donor-revenue model of 2019 tied a station’s dues to its aggregate membership revenue — not the revenue tagged to the programming that earned it. Moreover, by taxing donations, the model coupled NPR’s bill to a station’s fundraising success: raise more, owe more. NPR’s own chief operating officer would later call that “punished success.”

NPR’s most recent model reveals how far the system has moved away from the metrics that once aligned service, support and price. In 2025, Congress rescinded the $1.1 billion previously appropriated for CPB. This did more than shrink the funds available to purchase programming; the loss of CPB cut off the membership data NPR used to bill stations, forcing a new formula on short notice.

In broad strokes, the 2026 model prices dues on reach — the households inside a station’s broadcast contour. Reach is not service rendered or support received. A formula that does not know what programming is worth can offer a station no lever to pull — no way to enhance its public service, no way to increase its public support. Both returns are earned by serving significant audiences with significant programming — not by counting roofs under towers.

Reclaiming the measure

NPR may have adapted the best it could to the data, politics and other pressures of the day. Indeed, without a refreshed measure of listener-sensitive return per listener-hour at the program level, no national player can rebuild the virtuous dues model that once aligned network, station and listener interests so elegantly.

That is a real loss.

Stations, however, still have the metrics and tools to evaluate — and optimize — their own performance in public service and public support terms. Listener-hours can still be counted. Loyalty, appeal and pledge efficiency can still be analyzed at the program level. And at the enterprise level, so can public service (listener-hours), public support (membership and underwriting revenue) and the costs and returns of programming, development efforts and managerial overhead.

Strategic AudiGraphics was built for precisely this purpose: to quantify the local public service economy and lay it clearly on managers’ desks. A station may not be able to price a national program with the precision AUDIENCE 98 once made possible. But it can still ask the questions that matter most to its own future: How much public service are we providing? What does it cost? What support does it generate? Where are we strengthening the cycle? Where are we working against it?

That is where recovery begins — not with the resurrection of an old dues formula, but with the renewed discipline of measuring the worth of our endeavor by the service it actually provides and the support it actually earns.

This abandoned model and forgotten metrics can again demonstrate the dominance of our terrestrial service in the audio marketplace. Clear thinking, disciplined analysis and reliable data will not make hard decisions easy. But they will make those decisions better — and give our endeavor its best chance to remain valuable, viable and worthy of the trust listeners place in it.

David Giovannoni led the AUDIENCE 98 study and created AudiGraphics, in which he retains a financial interest. Its services, including the Strategic AudiGraphics referenced above, are offered and supported by Audience Research Analysis.