The Double Jeopardy Law: Why Smaller Local Competitors Always Look Less Loved

What the double jeopardy law actually says

The double jeopardy law describes a relationship between a brand’s market share and two separate things: how many people buy it, and how loyal those buyers are. The finding is that both move with share in the same direction. Smaller-share brands are bought by fewer people (the first jeopardy) and, among the people who do buy them, are bought slightly less often or held slightly less exclusively (the second jeopardy). A large brand enjoys the mirror image: more buyers, and marginally more loyal ones.

The name captures the unfairness of it. A small brand is not merely smaller; it is penalized a second time on a dimension, loyalty, that owners tend to treat as a matter of merit rather than of size. The pattern was first observed by William McPhee in 1963 in the context of media audiences and popularity, and was generalized to brand purchasing by Andrew Ehrenberg and colleagues, whose 1990 paper "Double Jeopardy Revisited" is the standard reference.

A pattern that keeps replicating

The reason double jeopardy is treated as a law rather than a study is that it has been observed again and again across very different markets. The original brand-purchasing work spanned packaged-goods categories; later replications extended it to services such as banking and insurance, and more recent analyses have found the same shape in newer categories including music streaming and ride-share. It is regarded within marketing science as one of the most replicated empirical regularities the discipline has produced.

This replication record is exactly why the correct framing is "a statistical law, not a moral failing." When a pattern holds across grocery brands, retail banks, insurers, and streaming apps, the explanation cannot be that all the smaller players in every one of those categories happen to be worse. The pattern is a property of how share and buying behavior relate, and it applies before any question of quality is even asked.

Why smaller businesses have fewer reviews, and slightly lower loyalty

Translate double jeopardy into the language of a local market and it starts to explain the review gap owners fixate on. A larger competitor serves more customers, so it accumulates more reviews simply as a by-product of volume. That is the first jeopardy made visible: fewer buyers means fewer reviewers. The second jeopardy suggests that, on average, the smaller business’s customers may also review, return, and recommend at a marginally lower rate, again as a function of size rather than of the work.

Read this way, "the other guy has more reviews" decomposes into at least three components. Part of the gap is pure volume (more customers, more reviews). Part may be the small loyalty penalty the law predicts. And part could genuinely be a difference in how well each business earns and asks for reviews, which is the only part any operator can move. The mistake is to attribute the whole gap to the third component and conclude that buyers prefer the competitor on the merits.

The review-count extension is an inference, not a proven law

A clear line matters here. The double jeopardy law itself is established: it was derived from purchase data (penetration and repeat-buying rates), and that is what has been replicated for sixty years. Its application to online review counts and star ratings specifically is a reasonable but unproven inference. Review counts are a plausible proxy for buyer numbers, and reviewing behavior is a plausible proxy for a facet of loyalty, but the law was not built on review data and has not been formally validated against it.

The useful and defensible claim is the directional one: because a larger competitor has more customers, some of its review advantage is a mechanical consequence of size that would exist even if the two businesses were identical. The unsupported claim would be to attach a specific number to that effect, or to present the review-count application as if it carried the same evidential weight as the underlying purchase-behavior law. It does not, and any read of your reputation should say so.

How brands grow: acquire buyers, do not just chase loyalty

Double jeopardy does not sit alone. It is part of a wider body of empirical generalizations associated with the Ehrenberg-Bass Institute and popularized in Byron Sharp’s How Brands Grow, whose central argument is that brands grow primarily by acquiring more buyers rather than by deepening loyalty among the buyers they already have. Growth comes from mental availability (coming to mind easily and often in a buying situation) and physical availability (being easy to find and buy), not chiefly from persuading existing customers to love the brand harder.

For a local operator, that reframes the correct response to a review gap. If the gap is largely a penetration effect, the lever is penetration: becoming more findable and more frequently considered across the surfaces where buyers now decide, so that more genuine customers arrive and, in the ordinary course, more of them review. Pouring effort into extracting extra loyalty from a small existing base is fighting the exact margin the second jeopardy says is hardest to move.

One caution belongs here. These generalizations were established across large, multi-brand, multi-decade datasets, and their application to AI-mediated and answer-engine discovery specifically is new and, as yet, untested. The principle that visibility and easy findability drive buyer acquisition is well grounded; the claim that it behaves identically inside a generative answer is an extension we treat as emerging, not settled.

What this means for reading a competitor’s reviews

The practical upshot is a discipline for interpretation. When a rival shows more reviews or a marginally higher rating, the first question is not "what are they doing that we are not," but "how much of this gap is explained by their size." A larger, longer-established competitor should be expected to lead on review count by default. The signal worth acting on is not the raw gap; it is the residual, the part that remains after size is accounted for.

It also disarms a dangerous temptation. If an owner misreads the whole gap as a verdict on quality, the pressure to close it artificially, by buying reviews, incentivising only positive ones, or gating unhappy customers away from public platforms, rises sharply. That is precisely the conduct the law would tell you is unnecessary, and it is also unlawful. The FTC’s Endorsement Guides require that reviews reflect honest experience and that any material connection between a business and a reviewer be disclosed, and the framework is actively enforced. The correct response to a predictable, size-driven gap is not to fake the difference away.

Measuring reputation without the vanity trap

Double jeopardy is, at bottom, a lesson about which numbers to trust. A raw review count is close to a vanity metric: it moves with size, it is easy to compute, and it invites the wrong conclusion. This is a familiar failure mode in measurement, and other disciplines have already disciplined themselves against it. The public-relations industry’s Barcelona Principles, revised in 2020, formally reject vanity metrics in favor of outcome-based measurement that is transparent, consistent, and valid.

The same posture applies to reputation. The metric that matters is not "how many reviews does the competitor have" but "where does this business stand once size, recency, and genuine buyer sentiment are read together, and what part of any gap is actually movable." A reputation reading built that way tells an owner something they can act on. A reading built on raw counts mostly tells them how big the other business is, which they could have guessed.

The summary

The double jeopardy law is one of the most replicated findings in marketing science, and it explains a large share of why smaller local competitors look less loved: they have fewer buyers, and their buyers are on average marginally less loyal, both as predictable functions of share. The extension of that pattern to online review counts is a sensible inference rather than a proven law, and it should be presented as such. Read together, the law and its limits point the same way: judge a reputation gap by the part that remains after size is stripped out, grow by becoming more findable to real buyers, and never manufacture the difference the data says you were always going to see.