Proprietary Research

“in the spring of 1951 these scientists tested these new types of water soluble plant foods to see how fast they could reach into every root, every leaf, every stem, every tissue. A sprinkling was applied to the roots of ordinary plants. Exactly 60 minutes later the leaves were cut off and pressed against special, extra-sensitive film.”

A 1950s print ad for a water-soluble plant food that narrates the maker's own timed absorption experiment step by step.

Why it’s this techniqueThe copy manufactures its own evidence by staging an in-house experiment: it dates the work to 'the spring of 1951', credits 'these scientists', and details the apparatus of 'special, extra-sensitive film', so the claim arrives as a lab finding rather than a boast. The structural tell is the step-by-step protocol, application, timed interval, measurement, then the verdict cued by 'The results show', a sequence no testimonial or borrowed authority produces; the seller ran the test and narrates it. The repeated 60-minute figure could read as a numbers play, but the number only carries weight because the staged experiment generates it.

“Claudia helped scientists test Crest against the control toothpaste. Sure the Crest users didn't brush harder? Couldn't. Nobody knew which side they were on. For a year, Claudia used toothpaste with no name on it. Turned out I'd been using Crest, and my side had 49% fewer cavities.”

A 1960s TV spot in which a participant recounts the year-long blind cavity trial the brand ran and the result on her side of the test.

Why it’s this techniqueThe copy stages a study the brand itself commissioned: 'scientists test Crest against the control toothpaste', then walks the reader through its blinding protocol, 'Nobody knew which side they were on' and 'toothpaste with no name on it', before releasing the finding, '49% fewer cavities'. The structural tell is that method carries the persuasion: most of the passage defends the experiment's design against the reader's objection ('Sure the Crest users didn't brush harder? Couldn't.') rather than describing the product. The precise figure and Claudia's testimony are outputs of the move; the manufactured evidence event is what the copy is built around.

“We don't know of anyone else on or off Wall Street who does research like this. Here are just a few of the more than 150 companies we've recently visited”

A 2000s direct-mail letter for an investment newsletter that opens on the firm's one-of-a-kind company-visit research program.

Why it’s this techniqueThe copy claims sole possession of an investigative operation: 'We don't know of anyone else on or off Wall Street who does research like this' stakes exclusivity on the firm's own fact-finding, then substantiates it with countable labor, 'more than 150 companies we've recently visited'. The structural tell is that the evidence is work the seller performed firsthand, site visits it can enumerate, rather than borrowed credentials or third-party studies; the authority rests on data no competitor holds. Read quickly, it resembles a simple superiority boast, yet every operative word describes the research process itself, so the homegrown investigation is the asset the copy sells.

“Slack commissioned a global survey in partnership with Qualtrics. Total sample: 18,149 desk workers and executives across 9 countries, including the United States, Australia, France, Germany, the United Kingdom, India, Singapore, Japan, and South Korea.”

A methodology note from a workplace-trends report describing the global survey the software company commissioned.

Why it’s this techniqueThe copy manufactures its own evidence base: 'Slack commissioned a global survey in partnership with Qualtrics' announces that the brand itself funded and owns the data, then stacks methodological specifics, 'Total sample: 18,149 desk workers and executives across 9 countries', to make the finding feel audited rather than asserted. The structural tell is ownership plus rigor markers: the sponsor is named as the actor who 'commissioned' the study, and credibility is borrowed from a named research partner, a precise headcount, and an enumerated country list. This is not mere statistic quoting; the excerpt exists to establish that the brand generated the research it is about to cite.

“We've taken that data and fed it into Kong. Creating a large language model (LLM) on the best-performing ad copy and creatives from over $200m in ad spend.”

A web page for an ad-copy AI tool explaining the private agency performance data its model was trained on.

Why it’s this techniqueThe copy manufactures evidence no competitor can replicate by converting the firm's own performance archive into the product's foundation: 'We've taken that data and fed it into Kong', building a model on 'the best-performing ad copy and creatives from over $200m in ad spend'. The structural tell is ownership of the dataset itself; the numbers describe internal results only this company holds, not borrowed studies or third-party citations, and the pitch's force comes from that exclusivity of access. The founder's '$7.8 billion (with a B) in ROAS' reads as authority, yet it serves only to certify the dataset, which is what the copy is built around.