Primary demand advertising also sells your competitors' product

Primary demand advertising sells a whole category, not one brand. In the US, 21 federal research and promotion boards buy it, funded by assessments on the industries they promote.

21 federal checkoff boards fund ads for a whole commodity, from beef to Christmas trees. Those ads lift every seller in the category, so no click report can show what they earned. The boards prove their return with an economic model instead, and the law makes them publish the result.

Primary demand versus selective demand

Primary demand advertising promotes a product category, such as milk or beef, so that the whole market grows. Selective demand advertising promotes one brand, so that it takes share from the others.

US law draws the same line. It says a producer’s own advertising is “typically designed to increase the share of the market” held by that producer, while a generic program is designed to increase “the overall demand” for the commodity (7 U.S.C. 7401).

	Primary demand	Selective demand
Goal	A bigger category	A bigger share of the category
Who gains	Every seller in the category	The advertiser
What proves it worked	Category sales against a no-ads scenario	Brand sales and share, often attributed conversions

The famous campaigns are checkoff campaigns

Four well-known examples were paid for by commodity programs, not by one company.

Campaign	Owner	Launched (owner’s own date)	What the owner says
Pork. The Other White Meat.	National Pork Board	1987	Its goal was to increase demand for pork and dispel its reputation as a fatty protein
The Fabric Of Our Lives	Cotton Incorporated	1989	The company was founded in 1970 because synthetic fibres were taking cotton’s share
Beef. It’s What’s For Dinner.	Beef Checkoff, through NCBA	1992	Seen at launch as “a catalyst for pushing beef to the forefront of consumer advertising”
got milk?	California Milk Processor Board	Board since 1993 (no campaign year given)	—

“The Incredible Edible Egg” is not in the table. No page from the American Egg Board gives its launch year, and a date copied from trade press is worth less than an empty row.

The 21 boards, what each one charges, and what it reports back

Congress has authorized industry-funded research and promotion boards since 1966. Each one funds itself with an assessment on its commodity, and its mission is to “maintain and expand the markets” for it. Dairy Act assessments totalled $343.6 million in 2024.

Read on 12 October 2026 from each program’s USDA page. Every cell is in the sources below.

Commodity	Board	Assessment	Published return per $1
Beef	Cattlemen’s Beef Board	$1 per head, each time cattle are sold	$13.41 (2019–2023)
Blueberries	U.S. Highbush Blueberry Council	$18 per ton	—
Christmas trees	Real Christmas Tree Board	$0.15 per tree (initial rate)	—
Cotton	Cotton Board, via Cotton Incorporated	$1.00 per bale plus 0.5% of its value	—
Dairy	National Dairy Promotion & Research Board	15 cents per hundredweight	$5.93 (1995–2024)
Eggs	American Egg Board	10 cents per 30-dozen case	about $8.11 (no period given)
Fluid milk	MilkPEP	20 cents per hundredweight	$5.14 (1995–2024)
Hass avocados	Hass Avocado Board	2.5 cents per pound	$2.47 (2018–2022, all six avocado demand organisations)
Honey	National Honey Board	1.5 cents per pound	—
Lamb	American Lamb Board	$0.007 per pound; $0.42 per head for first handlers	—
Mangos	National Mango Board	three quarters of a cent per pound	—
Mushrooms	Mushroom Council	Set yearly, capped at 1 cent per pound	—
Peanuts	National Peanut Board	$3.55 or $1.25 per ton, by grade	—
Pecans	American Pecan Promotion Board	$0.02 per pound in-shell, $0.04 shelled (initial)	—
Popcorn	Popcorn Board	6 cents per hundredweight	—
Pork	National Pork Board	$0.35 per $100 of market value	—
Potatoes	Potatoes USA	3 cents per hundredweight	—
Softwood lumber	Softwood Lumber Board	41 cents per thousand board feet	—
Sorghum	United Sorghum Checkoff Program	0.6% of net market value (grain)	—
Soybeans	United Soybean Board	0.5% of net market price	$12.30 (calculated from 1991)
Watermelons	National Watermelon Promotion Board	3 cents per hundredweight each, producers and handlers	19 to 1 (2017–2021)

Fourteen of the 21 rows have a dash. The law requires every board to publish an evaluation, so a dash means we found no return figure on the AMS page, the board’s own site or a USDA report, not that no evaluation exists. One cell is not the AMS figure: the AMS mushroom page gives “0.0055 cents per pound”, which reads as a unit error, so the table uses the cap in the order itself.

Do not rank the last column. The evaluators, periods and methods differ, and the dairy figure measures producer profit. Each number answers its own board’s question.

How they prove it works without attribution

A primary demand ad has no click to follow. A person who sees “Beef. It’s What’s For Dinner.” buys beef in any store, from any brand. There is no click ID, no pixel and no conversion to credit.

So the law asks a different question. Every board must fund an independent evaluation at least every five years and publish it. The beef and dairy evaluations follow the same steps:

1. Model demand from aggregate data: prices, incomes, spend on promotion, sales.
2. Run the model again with promotion set to zero. USDA’s dairy report calls this the “no promotion” scenario.
3. The gap is the effect. For 1995 to 2024, fluid milk use would have averaged 166.60 pounds per person instead of 185.41. For beef over 2019 to 2023, domestic demand would have been 8.5% lower.

This is an incrementality estimate: what would have happened without the ads. It is the question attribution reports cannot answer.

The gain leaks, and even the evaluations say so. The beef study that finds $13.41 per dollar notes that “beef packers and retailers likely benefited” as well, and that those benefits were not measured. The people who paid are not the only ones who gained.

When one company should pay for the category

For a single company, the leak is the whole problem. A category ad lifts every seller. If the lift splits by market share, a company with 20% of the category keeps one in five of the sales the ad creates. The other four go to its competitors.

Primary demand advertising pays for one company when:

• It holds most of the category, so most of the lift comes back to it.
• The category itself is at risk. Cotton’s program exists because polyester was taking its market. Pork’s campaign set out to change a reputation for fat. Neither problem has a brand-level fix.
• It can measure the category, not its clicks. A geographic holdout test or a model with a no-ads scenario can see a category lift. A last-click report credits it to whatever the buyer touched last.

What to do differently

1. Judge category ads on category sales. If the category grew where the ads ran and not where they did not, the ads worked, whoever got the sale.
2. Plan the measurement before the spend. Set aside holdout regions, or record spend and sales by week and region for a model. A test designed after the campaign cannot run.
3. Report the share you keep, not the total lift. The evaluation’s headline return is for the whole category. If the lift splits by market share, your return is that figure times your share.
4. Keep category and brand budgets in separate lines. Mixed together, the category ad’s lift shows up as cheap brand conversions, and the next budget review moves money to the wrong line.

The checkoff boards measure their ads with a model of total spend against total sales, then a no-ads run of the same model. They never track a single person. That method has a name in performance marketing. When does it beat attribution for your own channels? Start with marketing mix modelling.