Chapter 27 — Further Reading

Sources are grouped by how confident we are in them, following the book's citation policy: Tier 1 are works, institutions, and public records we are confident exist and can stand behind; Tier 2 are real practices and findings whose exact publication or precise magnitude we have not pinned down; Tier 3 is constructed teaching material in this book, labeled where it appears.

Marketing is the chapter of this book with the worst source hygiene in its surrounding literature. More confidently repeated, less verifiable numbers circulate here than anywhere else in restaurant publishing. Read everything below — including the Tier 1 material — with the question this book keeps asking: how would anyone know that?

Tier 1 — Verified canonical

Google's own documentation for Business Profile and local search. Free, current, and the only authoritative description of what the product does. Google states publicly that local results are driven principally by relevance, distance, and prominence; it publishes its content policies, including prohibitions on fake and incentivized reviews. Because the product changes several times a year, the documentation is the source and any book, including this one, is a snapshot. Read the current version before you act.

The Federal Trade Commission — Endorsement Guides and the rule on consumer reviews and testimonials. The FTC's guidance on disclosing material connections between a business and an endorser is long-standing and directly relevant to the comp and influencer policy in §27.6. In 2024 the Commission finalized a rule addressing fake and deceptive reviews and testimonials. Both are on ftc.gov at no cost and are written for businesses rather than lawyers.

The CAN-SPAM Act and the FTC's compliance guide for business. The statutory requirements for commercial email — accurate headers, non-deceptive subject lines, a physical postal address, a working opt-out honored promptly — are short enough to read in one sitting. Penalties are assessed per message, which is worth knowing before you send to 2,400 people.

The Telephone Consumer Protection Act (TCPA) and the FCC rules under it. The governing framework for marketing text messages, including the prior-express-written-consent standard. Note that several states have their own statutes with private rights of action, and that carrier-level requirements (campaign registration, content rules) sit on top of the law without being law.

The Credit CARD Act of 2009, which restricts expiration dates and inactivity fees on store gift cards and gift certificates, and your state's unclaimed property (escheat) statute, which governs unredeemed balances. These interact and they differ enormously by state. Your accountant is the right reader for both.

Levitt v. Yelp! Inc., U.S. Court of Appeals for the Ninth Circuit, decided 2014. The published opinion is the primary source for Case Study 1 and is more interesting than the summaries of it. Read it for the reasoning about what a platform owes the businesses it ranks — which is, in the court's account, essentially nothing.

The New York Attorney General's "Operation Clean Turf," concluded September 2013. The public announcement documents an undercover investigation into astroturfing and the resulting agreements with nineteen firms. A useful corrective to the assumption that fake reviews are a marginal problem.

Danny Meyer, Setting the Table. Named again here rather than only in Chapter 23, because this chapter's central claim — that marketing buys the first cover and hospitality buys the other three — is Meyer's argument with a cost per cover attached to it. Read it for the mechanism, not for tactics.

Byers, Mitzenmacher, and Zervas, on Groupon deals and Yelp ratings (published circa 2012). The empirical basis for Case Study 2's reputational-cost argument. Worth reading for the methodology — in particular how the authors work through competing explanations for a correlation, which is exactly the discipline this chapter's §27.9 asks of an operator.

Tier 2 — Attributed, specifics unverified

The "one star equals X% of revenue" literature. You will encounter this claim constantly, attributed to a small number of academic papers on hotel and restaurant ratings. Some of that underlying work is real. What is not defensible is the leap from a coefficient estimated on one market and one platform in one period to a number an operator can apply to their own business. Treat any specific elasticity you are quoted as unverified and ask what population it was estimated on. This book uses the offset identity instead precisely because it does not require a source.

Restaurant marketing budgets of 2–4% of sales. A widely repeated planning convention that appears across trade press and consulting practice. It is a reasonable sanity check and a bad plan; §27.1 explains why. Bellwether's 1.5% is deliberately at the low end and is defended by the cost-per-cover arithmetic rather than by the convention.

Daily-deal revenue splits (roughly 50% off face value, with the platform retaining roughly half of what the consumer pays). This is how merchants generally described the arrangement during the 2009–2012 period and how it was widely reported. Terms were negotiated and varied by merchant size and category; treat the split as the typical shape rather than a contractual constant.

Social media engagement benchmarks, optimal posting times, and platform reach statistics. This book gives none, and recommends that you treat the ones you are given with active suspicion. The sources are almost invariably vendor blog posts describing samples that are never characterized. Where a platform publishes its own aggregate data, read that instead — and remember the publisher's interest.

Yelp's recommendation software and its effect on solicited reviews. Yelp describes the software publicly and operators report the effect consistently. What is not verifiable is how it works or how much genuine content it filters. Plan around the behavior, not around a theory of the mechanism.

Gift-card redemption rates and average overspend on redemption. The pattern — most cards redeem, redeeming guests spend above face value, a minority never redeem — is real and consistent across retail and hospitality reporting. The specific percentages vary by category, face value, and season, and the figures in §27.7 are modeled illustrations rather than industry rates.

Tier 3 — Illustrative / constructed

• Bellwether's entire marketing plan — the \$23,250 budget, its five purpose buckets, the pre-opening timeline, the channel table, and every cost per cover derived from them. Constructed, internally consistent, and not a real business.
• Figure 27.2, the Google Business Profile performance summary — a constructed teaching example built to be internally consistent, not an extract from any real profile.
• Figure 27.6, the monthly attribution report — the Bellwether plan, modeled. Every number foots; none of them was measured.
• The visit-count distribution in §27.7 — modeled to be consistent with the plan's 9,035 guests at 4.0 average visits. It is what such a distribution plausibly looks like, not what any restaurant's distribution is.
• The gift-card, email-list, event, and press-spike calculations — all modeled, with the assumptions stated in the text so you can substitute your own.
• All exercise and quiz figures — constructed for legibility.

A note on what to read next

For the channel questions this chapter deliberately deferred: Chapter 28 on off-premise, which takes the incrementality problem introduced in §27.9 and makes it worth six figures; Chapter 29 on events, which is where the \$3,600 event line in this budget grows into a revenue line; and Chapter 24 on revenue management, which owns the question of what actually fills a Tuesday.

For the argument underneath the whole chapter: Chapter 23, which derived the \$18.40, the \$220.80, and the frequency-versus-reach comparison this budget is built on. If any of the arithmetic here felt asserted rather than earned, that is where it was earned.