What Sushi Restaurants Actually Buy: Takeout Container Purchasing Research

What do long-term purchasing records reveal about how commercial sushi takeout-container assortments are built, concentrated, expanded, and reordered?

ERS analyzed longitudinal purchasing histories using consistent commercial container categories, making assortment behavior comparable across differing product and SKU identities. The unit of analysis is a physical restaurant or foodservice location; parent operators are tracked separately.

What the Purchasing History Reveals

• Four architectures describe 80.03% of mature locations.
• Standard Rectangular Sushi Tray is recurring at 84.04% of mature sushi-container-buying locations and at more than 90% of locations within every dominant architecture.
• Locations carrying Standard and Large Rectangular trays adopt Round Sushi Party Platter within 24 months at 2.07× the rate of the Standard-only rectangular cohort.

Commercial container assortments function as operating systems: a recurring core supports recognizable extensions as menu and service needs broaden. Category popularity describes individual formats; the purchasing history reveals how those formats work together and how the system changes over time.

Complete-container units

3.25B+

Physical locations

14,794

Parent operators

9,490

States represented

36

January 2002–September 2026 · Edition 2026.1. A complete base with its matching lid counts as one unit. The study contains 3,251,175,900 units across 2,152,978 orders; components and carryout bags are excluded.

Which combinations recur together, rather than simply appearing somewhere in the purchase history? Looking at an assortment as a system distinguishes an everyday rectangular ladder from one that also supports party formats or broader-menu containers.

Four architecture families account for 80.03% of 9,991 mature locations; the remaining 19.97% form a diverse long tail. Median mature assortment breadth is 3 recurring categories.

Within the four dominant architectures, Standard Rectangular Sushi Tray recurs at 93.8%, 97.6%, 95.5%, 90.7% of locations, respectively, in matrix order. The dominant systems differ mainly in what is built around Standard Rectangular, not in whether it is present.

Co-occurrence makes that distinction visible. Locations sharing the same rectangular backbone can combine it with a broader size ladder, Round Sushi Party Platter, or a different mix of meal and roll formats. A popularity ranking would merge these different systems into the same Standard count. Architecture names summarize statistical signatures, not mandatory bundles; individual members can differ from the representative category set.

The research normalizes products into categories for cross-product comparison. The Size & Shape Guide provides the separate framework for defining, measuring, and mapping those categories.

Explore current container formats represented in this research → For a representative current example of Narrow / Roll Sushi Tray, see ED-01. A normalized category can include multiple products.

Breadth counts the categories a location keeps in recurring use. Concentration asks a different question: how much purchased volume sits in the largest recurring streams? A wide assortment can still have a compact operational backbone.

Standard Rectangular Sushi Tray is recurring at 84.04% of 9,770 mature sushi-container-buying locations. Unlike the 9,991-location architecture cohort, this narrower cohort requires at least one recurring rectangular, Narrow / Roll, Square, or Round Party format at the latest observed month; it excludes mature locations recurring only in broader-menu categories.

Among mature sushi-container-buying locations, Standard Rectangular Sushi Tray recurs at 84.04%, compared with 39.11% for Compact Rectangular Sushi Tray and 30.39% for Large Rectangular Sushi Tray. Standard is structurally separated from the rest of the field, helping explain why it appears across all four dominant architectures. That makes it central to these purchasing systems, not universally “best” for every application.

Penetration measures how widely a category recurs across locations. Concentration measures how much of a location’s purchasing runs through its largest recurring categories. Assortment breadth counts the formats; neither breadth nor penetration tells us how evenly a restaurant divides its volume among them.

Restricting concentration to mature locations with at least four recurring complete-container categories prevents a high top-two or top-three share from being mechanically guaranteed by a narrow assortment. Even with that restriction, the median shares are 75.09% and 89.85%, respectively. A restaurant may carry several formats while most unit volume still runs through a small operating core; extensions add variety without necessarily adding equal demand.

What a location already carries is associated with what it adopts subsequently. These comparisons start from a recurring assortment, exclude locations that have already adopted the destination category, and follow eligible locations for the same 24 months.

Party Platters Enter More Often Alongside an Established Rectangular Ladder

Among eligible locations with Standard Rectangular Sushi Tray and Large Rectangular Sushi Tray recurring, 40.15% adopted Round Sushi Party Platter within 24 months, compared with 19.36% where Standard was the only recurring rectangular tier—a 2.07× relative likelihood.

The two analyses look in different directions. The forward comparison shows higher later Party adoption among locations with a broader rectangular system; the prior-state analysis shows that most actual first Party adoptions followed an already-recurring base of at least two rectangular tiers. Together they support Round Sushi Party Platter as an expansion layer around the rectangular core, rather than another interchangeable everyday size.

“Within 24 months” means adoption anywhere in the qualifying forward window, not necessarily the immediately next category adopted. These are observational associations, not evidence that adding Large causes Party adoption or that every restaurant follows the same stocking sequence.

Representative current product: ED-65 is an example of Round Sushi Party Platter, a normalized category that can contain multiple products.

Recurring use does not imply the same purchasing rhythm. The interval distributions provide a separate check on the structural distinction between a core and its less frequent extensions.

Across qualifying purchase histories, Standard Rectangular Sushi Tray has a median interval of 21 days, compared with 52 days for Round Sushi Party Platter. The distributions overlap; category identity does not determine an individual location’s schedule.

Across qualifying category purchase intervals, Standard Rectangular Sushi Tray has a median of 21 days and an interquartile range of 16–28 days; Round Sushi Party Platter has a median of 52 days and an interquartile range of 33–85 days. Party is both slower and more variable, consistent with a more conditional purchasing role. The overlap still matters: neither category has one fixed reorder cycle.

Specialized Does Not Necessarily Mean Infrequent

Rice / Entrée Bowl and Compact Rectangular Sushi Tray each have a median qualifying purchase interval of 24 days and an interquartile range of 17–33 days. The medians are 26 days for 5-Compartment Bento Box and 28 days for Narrow / Roll Sushi Tray. Rice / Entrée, Bento, and Narrow / Roll are less prevalent across locations than Standard, yet their purchasing cadence can resemble regular operating formats where they are used.

Category role has two dimensions: how widely a format recurs across restaurants and how frequently it is purchased at locations that use it. A format can be specialized across the population without being occasional within its own purchasing histories. Assortment structure and reorder cadence are different views of the data, but they produce a compatible picture of core formats and conditional extensions.

These intervals measure purchasing, not consumption, depletion, lead time, inventory coverage, or recommended reorder timing. Population-level patterns do not determine an individual restaurant’s assortment. The Assortment & Reorder Planner connects the framework to restaurant-specific menu, usage, case-pack, inventory, and replenishment inputs.

What the Four Exhibits Add Up To

• Assortments are structured. Recurring combinations resolve into a few common systems alongside a meaningful long tail.
• Dominant systems share a backbone. Standard Rectangular spans the families, while most volume runs through a small operating core.
• Expansion depends on the existing system. Different recurring starting assortments are associated with different subsequent adoption rates.
• Category role appears again in purchasing cadence. Frequent core purchases coexist with slower, more variable extensions—and specialized formats can remain regular within their own users.

Commercial container purchasing is not well described by a single popularity ranking. Restaurants build systems with a core, recognizable branches, and purchasing rhythms that reflect the role each format plays.

ERS analyzed category-normalized order lines across physical locations and parent operators. Differing product/SKU identities are compared through consistent commercial container categories. The analysis reconstructs month-end recurring assortments from dated qualifying purchases and compares those states with subsequent category adoption.

Qualifying adoption

The first consolidated category purchase of at least 100 complete units. Smaller lines remain in volume totals but do not establish adoption or recurrence.

Recurring category

At least 3 qualifying category purchase events in the trailing 12 calendar months, evaluated at month-end.

Mature location

At least 24 calendar months between first and latest qualifying purchases, at least 18 qualifying category events overall, at least 12 in the trailing year, and a nonempty recurring assortment.

Observation window

January 1, 2002–September 27, 2026. Architecture and penetration use each location’s latest qualifying observed month. Adoption comparisons require a complete 24-month forward window and activity in at least 18 of those months.

Reorder interval

Elapsed calendar days between consecutive qualifying purchases of the same category at the same physical location, after consolidating purchases within anchored two-day distributor visits.