Topic
Measurement
Before tactics, before budget, before creative: can this account tell what is working? In most small accounts the honest answer is no, and everything built on top of that answer inherits the problem.
FrameworksThe margin-back budgetA budgeting method that derives an SMB's marketing spend backwards from its own prices, margins and capacity, where contribution per customer sets what a lead may cost and capacity sets how many leads to buy, instead of picking a percentage of revenue that fits no business in particular.
FrameworksAARRRA measurement structure that splits the customer journey into acquisition, activation, retention, referral, and revenue, and assigns exactly one metric to each stage so no stage can borrow success from another.
FrameworksSee-Think-Do-CareAn audience framework that segments by purchase intent instead of demographics and requires a separate message, channel, and success metric for each intent stage, because judging awareness activity by conversion metrics kills it every time.
FrameworksThe holdout testThe incrementality method that measures what advertising actually causes by withholding it from a comparable region or audience and counting the conversions that happen anyway, because platform-reported conversions include sales that would have occurred regardless.
FrameworksThe 95:5 ruleJohn Dawes's estimate, published with the Ehrenberg-Bass Institute and the LinkedIn B2B Institute, that only about 5% of B2B buyers are in the market in any given quarter, which makes advertising's main job building memory with the 95% who will buy later, not converting the few buying now; in any category the share follows from how often it is bought.
TakesGoogle Ads AI Max migration: a small-account checklist before 1 OctoberGoogle is moving campaign-level broad match and Automatically Created Assets into AI Max in September, in place and with no way back. Neither Google's +7% nor smec's 35% lower ROAS measures what that does to your account: a small account should narrow its scope, check five settings, and let a holdout on its own bookings decide.
TakesHelloFresh failed the market test at €7 billion, then passed itBernstein put HelloFresh's lifetime value to acquisition cost at 0.8x in 2022: it was paying more for a customer than the customer paid back. That is the margin-back budget's failure condition, at €7 billion of revenue rather than €180 a contract. The correction was the model's own prescription: fix the inputs before buying more leads.
TakesQuibi had no bad case at allQuibi raised $1.75 billion, projected more than 7 million subscribers in year one, and shut down after six months with about 500,000. The detail that matters is a word: in investor presentations, 20 million subscribers was the base case. When the base case is the good case, the bad case is missing by construction.
TakesPeloton published the good case and called it the planIn August 2021 Peloton guided $5.4 billion, no range, no bad case. It cut in November, cut again in February, and closed the year at $3.58 billion, 34% below the first number. Each cut was received as a betrayal because no scenario existed for it to land in. The same equation runs inside a company, pointed at its own forecasts.
TakesCyberpunk 2077 shows what expectation debt costs, itemisedEight million pre-orders covered the whole cost of the game before a single review. Eight days after launch Sony delisted it and refunded buyers, and the shares fell 43% from their peak. Expectation debt is the only zone where a product is a commercial success on day one and a liability by day eight.
TakesDisney runs the beatable band as a daily operating systemJoin a 60-minute queue at Walt Disney World and you will very likely be out in 40. Independent measurement puts actual waits at 55 to 71% of posted times, averaged across the four parks. That is a promise set at bad-week delivery, and a surplus handed to every guest at a cost of zero.
TakesDomino's lowered the bar, then cleared itIn December 2009 Domino's told the public its pizza was bad, and same-store sales rose 14.7% the next quarter. Read through one equation, satisfaction equals delivery minus expectation, the campaign was not honesty for its own sake. It moved the promise below the pizza.
TakesExpectation management case studies: six brands, one equationThe most-cited turnarounds and collapses of the last fifteen years were expectation events, not product events. Six cases with public numbers show the same arithmetic that runs a €2,000-a-month campaign for a dental clinic.