📢 How Advertising Works

What you’ll learn

1. What You Are Actually Looking AtReplace the 'ad as message' model with the 'ad as purchased attention' model, and meet the through-line.An advertisement is the delivery of a transaction: a company that wanted your attention bought some from an organisation that had assembled it. So ad-funded media are attention wholesalers — content is the bait, the audience is the inventory, the advertiser is the customer. The course's argument: attention is bought and sold like a commodity, and the buyer usually cannot tell what they bought.
2. The Old Machinery: Reach, Frequency, and a Price per ThousandLearn the pre-digital vocabulary — reach, frequency, CPM — and why wastage was structural.Before targeting you could not know who was looking, so attention was bought in bulk: reach × frequency = impressions, priced per mille. Frequency is where budgets leak, because in a mass medium you buy slots rather than people and the same heavy viewers keep reappearing. The 'half my advertising is wasted' line — attributed to Wanamaker without reliable sourcing — names a measurement problem, not a claim that ads don't work.
3. How the Ad Knows About YouUnderstand targeting as buying people rather than contexts, and the four ordinary ways an ad finds you.The structural shift was from buying a context and hoping the right people were in it, to buying the person wherever they are — which is why data about you became the product. Four routes do the work: your behaviour on a site (retargeting), behaviour across sites, data you handed over, and statistical resemblance (lookalikes). The FTC's report lists six interface designs that obscure or subvert privacy choices, which together describe the average cookie banner.
4. The Auction You Never SeeFollow the real-time auction that prices a specific person's attention in under 100 milliseconds.Between your tap and the page, a bid request fans out to bidders who value you rather than the slot, and the auction clears. Second-price rules make honest bidding optimal; the industry's move to first-price put strategy back in, spawning bid-shading. And in search, ads are ranked by something like bid × predicted quality — so what you see is the winner of a contest whose formula belongs to the platform.
5. Measurement: The Industry's Open SecretSee why accurate counting does not answer advertising's only real question, and what does.When eBay switched off brand-keyword search ads, about 99.5% of the forgone clicks were immediately recaptured by free natural search — the ads had been buying visitors eBay already had. The same paper measured eBay's non-brand search ROI at over 4,100% by ordinary regression, over 1,400% with controls, and −63% by experiment. Last-click attribution compounds this by handing credit to whatever ad stood closest to the purchase; only incrementality experiments answer the question, and nobody in the chain is paid by that answer.
6. The Edge, and Who Polices ItName the deceptive edge of the industry and the structural pressure that produces it.The persistent temptation is for an ad to stop looking like an ad; the FTC's policy statement treats disguised advertising as deceptive when it misleads consumers into believing it is independent or impartial. The Credit Karma complaint shows the mechanism: an A/B test found an allegedly false 'pre-approved' claim outperformed a truthful one, and that version shipped — because response is measurable and honesty is not. The honest defence also stands: advertising funds much of human communication, and without it incumbency would be nearly unassailable.
7. A Field Guide to the Ads Around YouLeave with four questions to ask of any ad, and an accurate model of what an ad is.Why me (retargeting, a lookalike, or data you gave)? What did this cost and who decided (an auction that priced you specifically)? How will they know it worked (last click, which credits the ad that changed least)? Is it wearing a uniform? And the hopeful half of the eBay result: the ads that feel most invasive are largely the ones standing in a doorway you were already walking through.

Questions this course answers

Grounded in trusted sources

• Tom Blake, Chris Nosko & Steven Tadelis, 'Consumer Heterogeneity and Paid Search Effectiveness: A Large Scale Field Experiment', Econometrica 83(1), 2015, 155–174 — https://onlinelibrary.wiley.com/doi/abs/10.3982/ECTA12423 (working paper: https://faculty.haas.berkeley.edu/stadelis/BNT_ECMA_rev.pdf)
• US Federal Trade Commission — 'Bringing Dark Patterns to Light', staff report, September 2022 — https://www.ftc.gov/reports/bringing-dark-patterns-light
• US Federal Trade Commission — Enforcement Policy Statement on Deceptively Formatted Advertisements (2015)
• US Federal Trade Commission — Restore Online Shoppers' Confidence Act (ROSCA) and the Negative Option Rule
• Encyclopaedia Britannica — 'Advertising'

Every Wunder lesson is built from real, reputable sources — never invented.

Related Math courses