Anyone can make an ad now. Knowing which one will sell is the skill. Here is the evidence that judging the work well moves sales, share and profit, and why the Academy is built to train it.
Every number on this page names its source
Media buys the chance to be seen. The work decides what that chance is worth. Four different kinds of study point the same way.
of the sales lift advertising produces comes from the creative itself. That is more than reach, targeting and the brand put together.
Industry study · Source 1
the sales return on each euro for the most creative campaigns, in a study that matched consumer ratings of 437 campaigns to real purchases.
Field study · Source 3
the media money a dull ad needs to do the job of an interesting one.
Vendor model · Source 4
of advertising’s profit arrives after the first three months, long after a click report has closed.
Industry model · Source 5
A machine can draft a hundred ads before lunch. So can everyone you compete with. When making is nearly free, the advantage goes to whoever can tell which ad deserves to run.
lower scores for ads made only with AI, on short-term sales potential and long-term brand equity. Researchers remade 20 ads from 10 major brands and tested both versions on 3,000 people. The AI versions fell short most on story, emotion and point of view, though only 13% of viewers felt sure they could spot them.
Controlled test · Source 6
comparisons in which AI help made a group’s ideas less varied. Brands are remembered for being different, so sameness is a cost.
Peer-reviewed · Source 7
None of this says AI cannot make good work. It says someone still has to know good work when they see it.
In a study of 339 professionals scored against 13,248 audience ratings, the people who make the work predicted which new ideas would succeed better than managers did.
Peer-reviewed · Source 8
A 2023 review of creativity training for adults found that teaching people to evaluate and choose ideas worked better than teaching them to produce more of them.
Peer-reviewed · Source 9
Seeing strong examples early, and again between drafts, made people’s work more creative. Seeing them only at the end mostly produced copies.
Peer-reviewed · Source 10
The gains are real and modest. Once publication bias is corrected, creativity training moves people by about a third of a standard deviation. That is why the Academy pairs every master class with exercises marked against a rubric built with the Legends.
Peer-reviewed · Source 11
Creativity pays under conditions. These six have the strongest evidence behind them. Ask them of every idea, yours included.
The memory has to land on the brand. Ads with strong brand assets were 2.5 times more effective than ads with weak ones.
Vendor study · Source 12
Originality alone does not sell. Across 878 measured effects, creativity worked when the idea was both new and relevant to the brand.
Peer-reviewed · Source 13
Emotional campaigns were about twice as likely as rational ones to deliver top profit growth over the long run.
Industry databank · Source 14
Campaigns with a recurring character or device were 37% more likely to grow market share.
Vendor study · Source 15
Budget, time and channels track effectiveness closely. Brilliant work on a thin, short burst rarely shows up in the numbers.
Industry databank · Source 16
About 8% of US internet users made 85% of display-ad clicks. Judge the work on share, price and profit.
Industry study · Source 17
Credit in advertising is shared. Each of these was a team's work, made at an agency a Legend led.
Chuck Porter
Florida’s truth anti-smoking campaign came out of Crispin Porter + Bogusky, Chuck Porter’s agency, in 1998. The national truth campaign that followed is credited with about 22% of the fall in US youth smoking from 1999 to 2002, roughly 300,000 fewer young smokers.
Peer-reviewed · Source 18
Tim McClure
Tim McClure and Mike Blair of GSD&M created it in 1986. The Texas Department of Transportation credits it with cutting roadside litter by about 72% by 1990.
Official figure · Source 19
Lee Clow, Chuck Porter, Susan Credle
The Effie is the industry’s award for results. Agencies led by three of the faculty took its top prize: iPod Silhouettes at TBWA\Chiat\Day (2005), Whopper Freakout at Crispin Porter + Bogusky (2009) and Whopper Detour at FCB New York (2020).
Know what kind of evidence you are reading. Much of this research comes from companies that measure advertising for a living, and much of it is a correlation or a model rather than an experiment. Every source below says what kind it is, so you can weigh it.
Creativity is no guarantee. In the IPA’s databank, award-winning campaigns once grew market share up to 12 times more efficiently than the rest. By 2018 that edge was nearly gone, as awards rewarded short bursts and budgets thinned.
Industry databank · Source 23
The lesson is not “be more creative.” It is “judge the work by what it does.” That is what the Academy teaches.
Yes, on the best evidence available. In NCSolutions’ studies of about 450 consumer-goods campaigns, the creative explained about half of the sales lift advertising produced, more than reach, targeting or the brand itself. The effect depends on the work being branded, relevant, sustained and funded.
Yes, with modest gains. Reviews of creativity training find that teaching people to evaluate and choose ideas works better than drilling them to generate more, and that studying strong examples early, before and between drafts, helps.
Not yet, on average. In a 2026 test of 20 ads on 3,000 people, ads made only with AI scored 14 to 17% lower than the human-made originals on predicted sales and brand equity. AI-assisted work chosen by people with strong judgment can test very well, which is the point.
Industry model
Industry databank
Industry databank
Industry study
Industry databank
Every master class shows how a Legend decided, then asks you to decide.