Section 1
The ensemble: four lenses on one market
Model 1: Market structure (why the supply exists at all) Start with the economics, because the whole thing is a market and markets exist when supply, demand, and a price clear. On the supply side, producing a fake review is cheap, scalable, and low-risk: accounts are plentiful, text is trivial to generate, and prosecution is rare. On the demand side, buyers value the outcome highly, because a rating now gates discovery in the map pack and in AI answers, so moving a competitor's average or lifting your own has real revenue attached. Cheap to produce, valuable to buy, hard to police: that is the exact profile of a market that will exist and grow. The structural read gives you the first defensive insight. You cannot make the market disappear, any more than a shop can abolish shoplifting. What you can do is raise the cost or lower the payoff of attacking you specifically, so the supplier chooses an easier target. Every effective defense in this piece is a version of that: make yourself the more expensive, less rewarding victim. Model 2: Game theory (the attacker types and their payoffs) Market structure tells you a market exists. Game theory tells you it is not one player, it is several, each running a different game with a different payoff. This is the core of the piece, because the attacker's game determines your defense. The competitor saboteur. Payoff: relative advantage. They do not need to destroy you, only to sit above you in the ranked list a buyer sees. Their game is quiet and patient: a slow trickle of plausible one-stars, sometimes with specific-sounding but false detail, timed to keep your average just below theirs. Because the goal is relative position, the attack is often small and sustained rather than a dramatic flood, which is precisely why it is easy to miss and hard to prove. The extortion crew. Payoff: a direct payment from you. Their game is a scripted funnel: a burst of one-stars as proof of force, a demand, a deadline, an escalation ladder if you engage. The attack is deliberately visible and frightening, because fear is the conversion mechanism. Unlike the saboteur, the extortionist wants you to notice, and wants you to reply. Their whole model depends on your reaction. The review brokerage. Payoff: recurring revenue from clients who buy reviews (usually positive ones for themselves, sometimes negative ones against rivals). This is the industrial layer: networks that sell reviews at scale as a service. You may encounter them as the production engine behind a competitor's positive-review inflation that makes your honest rating look weak by comparison, or as the supplier a saboteur hires. Their game is volume and evasion of platform detection, played across thousands of businesses, not personal. The game-theoretic point is that these are different players optimizing different things, so a defense that changes one player's math may do nothing to another's. Refusing to pay defeats the extortionist's game entirely and does nothing to the saboteur, whose payoff never depended on your money. A large review base that absorbs shocks blunts the saboteur and the extortionist alike but does little about a brokerage inflating a competitor. You have to know which game is being played. Model 3: Network and information economics (why attacking you pays now more than before) Why has this market grown? Because the value of the thing being attacked has risen. A review used to be a trust signal read by a human who had already found you. It is now also a ranking input: rating feeds the map pack and the shortlists that AI answer engines assemble, so it partly determines who gets found at all. When rating gates discovery, moving a rating moves demand, and moving demand is worth paying for. The network structure of local search, a few visible slots that a rating helps allocate, is what converts a fake review from a mild annoyance into a lever on revenue. That is the demand-side engine behind Model 1's market. This lens also explains the target selection. Businesses whose rating sits near a rounding edge (a 4.4 that shows as 4 versus a competitor's 4.5) are high-value targets, because a few fake reviews there tip a visible threshold. A business at 4.9 with two thousand reviews is a low-value target, because no realistic number of fakes moves the displayed number. The network economics literally price your attractiveness as a victim. Model 4: Behavioral economics (why the attacks work on you) The final lens is about the target's psychology, because the attacks are engineered around predictable human reactions. Owners over-weight the star average and the most recent negative review, feel each fake as a personal wound, and react fast and emotionally, which is exactly the behavior the extortion funnel monetizes and the saboteur relies on to keep you distracted. The behavioral read is uncomfortable but useful: your instinct to reply angrily, to pay to make it stop, to solicit a defensive burst of five-stars, is the reaction the attacker's model assumes. Defusing your own reaction is itself a defense, because it denies the extortionist the conversion and denies the saboteur the panic that leads you into filter-tripping mistakes.
Section 2
The combined read: match the attacker to the defense
Put the four models together and the picture resolves. A fake-review market exists because supply is cheap and demand is valuable (Model 1); it is populated by distinct players with distinct payoffs (Model 2); their payoffs rose because rating now gates discovery (Model 3); and their tactics are tuned to your predictable reactions (Model 4). The defensive implication is a diagnosis-first response: identify the attacker before you act, then apply the defense that changes that specific attacker's math.
Section 3
The proposed response: a ranked, portfolioed plan
Diagnosis is worthless without a sequence. Here is the response ordered by speed and reversibility, so you know what to do now, what to hold as a hedge, and what to defer. Do now, regardless of attacker (the shock absorber). Build and maintain a large base of genuine, recent, specific reviews. This is the one defense that helps against every attacker in the table, because it lowers the payoff of attacking you at all: when your rating cannot be moved by a realistic number of fakes, the saboteur gets no relative gain, the extortionist has no leverage to sell, and a brokerage-inflated rival no longer makes your honest number look weak. It is slow to build, which is exactly why it must start before you are attacked. Velocity and recency are the balance sheet you draw down during a crisis. Do now, when attacked (the incident response). Diagnose the attacker from the signature, capture evidence immediately (screenshots, timestamps, reviewer histories, any demand message), and report the cluster as coordinated rather than reporting reviews one at a time. If a payment demand is attached, do not reply and do not pay, and file with the platform's abuse channel and the relevant fraud body. Hold public replies to a single calm statement, because the reply audience is the next prospect, not the attacker. Hedge (build the optionality). Reduce your dependence on any single platform's rating by earning proof across the surfaces your buyers actually use and owning a direct enquiry route, so a successful attack on one profile does not sever your demand. This is insurance: you hope not to need it, and you are glad to have it when a rating drop lands. Defer, with a trigger. Legal action against a defamatory reviewer, or a formal law-enforcement pursuit, is slow and costly and rarely the first move. Defer it, but set the trigger clearly: a court order finding specific statements defamatory is the lever that works when policy-based removal fails, and a documented, sizable extortion demand is the point at which a fraud complaint is worth filing. Prepare the evidence now so the option is live if the trigger fires. The base-rate check. Before assuming every bad review is an attack, run the honest test: is this a coordinated cluster of low-history accounts with no specific detail, or is it real customers describing a real failure? The base rate of "genuine unhappy customer" is higher than wounded owners want to believe, and misclassifying a real complaint as an attack, then reporting it as fake, damages your credibility and your ability to fix the actual problem. Diagnosis includes ruling out that the review is true.
Section 4
The transparency card and the blind spot
Each model in the ensemble earns its place by what it reveals, and each has a boundary. Market structure explains why the supply exists but says nothing about who specifically is targeting you. Game theory sorts the attackers by payoff but assumes they are rational and distinct, when a real attack can blur types (an extortionist who is also a competitor). Network economics explains why attacks pay now but treats platform behavior as fixed, when a platform rule change could reprice the whole game overnight. Behavioral economics explains your reactions but risks over-attributing strategy to what is sometimes just one angry real customer. Here is what no model in this ensemble sees: the platform's private detection state. Every defense above assumes you are managing your side of a system whose actual decision logic, removal thresholds, filter behavior, and appeal criteria, is undisclosed, changes without notice, and is not something you can observe or appeal into. Google and its peers publish enormous aggregate removal numbers and almost no per-case data, so every hit-rate and every "this reporting method works better" claim in this space, including the directional mappings here, is inferred from stated policy and practitioner observation, not measured. The single largest actor in the fake-review economy, the platform that both hosts the market and polices it, is the one whose behavior you can least predict. Build your defenses to be robust to that uncertainty: prefer the moves that help regardless of what the platform's black box decides (a large genuine review base, owned demand, calm incident response) over the moves that bet on a specific removal outcome you cannot control. The fitness test: You understand the fake-review economy if, when an attack lands, your first act is to diagnose the attacker (saboteur, extortionist, or brokerage) from its signature before you respond, if your standing defense is a genuine review base large enough that no realistic number of fakes moves your displayed rating, and if your demand does not depend entirely on one platform's rating surviving. If your instinct is still to treat each bad review as random bad luck, reply in anger, or pay to make it stop, you are reacting exactly as the market's suppliers have priced you to react, and you are the target their model was built for.