The lack of data proving the effectiveness of ad campaigns

Advertising now possesses astonishing machinery for making mediocrity punctual. Audiences can be sliced into exquisite segments. Thirty-seven formats can appear almost instantly. Reporting can describe each movement in reassuring detail. Meanwhile the central thought—the bit a human might notice or remember—receives whatever time remains after the apparatus has been admired.

Data and creativity are not rivals in this mess. Evidence can reveal a barrier, distinguish a recognised asset from an internal favourite and test whether behaviour changed. The stupidity is economic: expensive precision surrounds a cheap assumption. Twelve carefully modelled audiences receive twelve versions of the same beige proposition. When an unusual idea finally appears, it threatens production decisions already made and measures already chosen. The ordinary idea has acquired institutional support simply by arriving before anyone expected originality. This creates a wonderfully professional failure: the work is delivered everywhere, on time, with proof that it was seen, and leaves almost no trace in memory. An agency can spend days debating whether an audience is “urban progressive” or “connected achiever”, then give both the same generic reason to care. The segmentation looks clever because the weakness sits one layer earlier.

A genuinely central idea changes the rest of the plan. It may alter where the work belongs, which formats deserve making and what evidence would count. That makes early exploration look inefficient beside immediate production, because thought refuses to generate reassuring quantities on demand. The strange option often looks unfinished precisely when the safe one looks ready for production. That asymmetry rewards familiarity before either idea meets an audience. The commercial question is less flattering than the process chart: how much effort increased the chance of being remembered, and how much documented the distribution of something forgettable? Perfect delivery cannot rescue a proposition with no pulse. It only makes the post-mortem easier to schedule.

Behavioural principles

Behavioural ideas at play in this post

Short, plain-English explanations of the principles behind this post, with links to related books and examples in the archive.