How to Compare Two Fast-Food Chains Without Fooling Yourself

The honest comparison between two fast-food chains is narrower than most people want it to be. It holds inside one city, county or state at a time, over one period, and only when the two chains’ margins of error do not overlap. There is no single number that says which of two chains is cleaner, and this site will not print one.

That is not modesty. It is what the records can carry.

Why a comparison cannot cross places

Each place on this site counts a different event. Florida counts whether an inspection recorded a high-priority violation. New York City counts whether a first visit scored 14 or more points. Chicago counts a Fail; King County counts an Unsatisfactory; Clark County counts a 5-demerit violation; Austin counts a score under 90; New York State counts a red violation.

Those are not the same measurement in different cities. They are different measurements, written by different agencies, against different rulebooks, with different thresholds. Averaging a chain’s Florida rate with its Chicago rate produces a number that describes nothing that exists. Ordering two places by their rates tells you which agency writes more findings down, not which city’s restaurants are cleaner.

So the compare tool runs each place separately and shows you the results side by side. The only thing it adds up is places — how many of the 7 we cover put one chain clearly below the other. A count of places is not arithmetic on rates, and it is the one cross-place statement this site makes.

What an overlap means

Every figure here ships with a 95% range: the band of values consistent with the inspections on file. A chain measured on 150 inspections has a wide band; one measured on 3,000 has a narrow one.

When two chains’ bands overlap in a place, this site calls it a tie, whatever the two midpoints say. One chain showing 18% against another’s 22% looks like a difference, but if the first could plausibly be anywhere from 14% to 23% and the second from 17% to 28%, the records have not separated them. The next year of inspections could reverse the order without anything changing in either kitchen.

The rule the tool applies is deliberately strict: two chains are apart only when their ranges do not touch at all. A formal test on the difference would separate a few more pairs than that. We use the stricter rule because it is one a reader can check by looking at the bars — the same rule the charts teach everywhere else on the site, and a rule that errs toward saying “too close to call.”

A tie is a result

Most comparisons on this site come back overlapping. That is the honest finding for two brands running similar operations under the same local rules, and it is worth more than a manufactured ranking.

It is also why nothing here is numbered. There is no “#1 cleanest chain,” no top ten, no league table. Ordinal rankings imply that adjacent positions are real differences, and in this data they usually are not. The chains are drawn in rate order for legibility, and that order is not a ranking.

What a comparison still cannot tell you

• Which chain is better run. Chains differ in what they handle. A coffee counter runs fewer cooking, cooling and raw-handling steps than a kitchen frying chicken, so part of any gap is the menu rather than the management. A fair comparison of two chains with very different food is not really a comparison of two operators.
• Anything about one restaurant. Every figure sums up many inspections at many locations. The chain that comes out lower in a place still has locations that failed inspections there, and the higher one has locations with clean records.
• Whether a chain has improved. The figures cover a multi-year window as a block. They are not a trend line.
• Chains with too few inspections. A chain needs 100 counted routine inspections in a place before it gets a range there. Below that it is listed with its count and left uncharted, and the tool has nothing to compare.

Using the tool

Pick two chains. The tool finds every place where both clear the minimum, compares them there, and tells you how many places separated them and how many came out tied. A comparison is linkable, so a result can be cited directly.

If two chains share no place in common, there is no comparison to make, and the tool says so rather than reaching for a national number.

Sources

• The comparison rule, the minimum sample and the interval method are set out on this site’s methodology page, which also names every government dataset behind the figures.
• Restaurant names are matched to chains with a published, versioned table: chain names lists every pattern and every name it matched.

Informational only. This is a reading of public inspection records, not food-safety, legal or dining advice, and no figure here is a grade of any individual restaurant. A figure describes what inspectors recorded across a set of visits in one place over one period — not a brand, and not the location near you.

Last reviewed: September 2026 · Against this site's own published method.