How Large Is the Average Delivery Menu?

How Large Is the Average Delivery Menu?

A typical delivery menu in DealMeal's DC-area study contained 67 listings. That is the median, meaning the middle menu when menus are ranked by size. It describes a typical restaurant listing more usefully than an arithmetic average influenced by unusually long menus.

DealMeal Labs analyzed the latest recorded menu observation for each platform-restaurant, retaining 4,926 menus with 428,717 listings after excluding flagged shops. The records cover September 7–October 12, 2026. The findings show how menus differ across apps, how larger menus are organized, and what menu length can tell us about finding promotions.

Quick answer

The median delivery menu contained 67 listings; the middle half ranged from 37 to 114. The median recorded section count was 10. These are DC-area findings, with a shop filter that misses some storefronts, rather than a national benchmark.

Counting each listing equally gives a different perspective: the median item belonged to a menu with 122 listings. Longer menus account for more of the catalogue. Among 1,482 matched stores that had promotions, menu length showed no meaningful association with the number of distinct offer titles. The study did not measure scrolling, taps, or discovery time.

Key findings

• Typical size: 67 listings per menu, with 37–114 covering the middle half of the 4,926-menu study population.

• Catalogue concentration: the quarter of menus above 114 listings held 53.4% of all retained listings.

• Menu structure: larger menus had both more sections and more listings within each section.

• Same branch, different menu: DoorDash had a median paired advantage of 36 listings over Grubhub; its 5-listing advantage over Uber Eats fell to zero at the median when only dishes were counted.

• Promotion count: longer menus did not predict more distinct offers among the stores that ran promotions in the matched sample.

What counts as a delivery menu item?

Menu size here means the number of listing rows in one recorded menu observation. It includes dishes, drinks, combos, retail products, and modifiers. A listing is a choice presented in the menu; it is not necessarily a unique meal.

If the same drink appears under separate sections, both rows count. Likewise, the same physical restaurant on different apps contributes a separate menu for each app. The 4,926 retained menus represent 3,691 physical kitchen branches, rather than 4,926 different locations.

The median dish count was 46, the median drink count was 8, and the median combo count was 1. Those are separate medians, so adding them does not reconstruct the median total menu. This distinction matters when comparing food choice with the length of a menu listing.

What the delivery menu size distribution looks like

There is a broad range around the typical menu. The lower quartile was 37 listings and the upper quartile was 114; the 90th percentile reached 179. A restaurant menu with more than a hundred listings was therefore common enough to matter, even though it was larger than the median.

Menus with 100 or more listings represented about 31% of the retained menus but held 61.0% of their listings. That threshold differs from the quarter above 114, which held 53.4%. Each percentage answers a different question about concentration.

DealMeal data insight: a typical menu and a typical listing differ

Give every restaurant menu equal weight and the median is 67 listings. Give every item listing equal weight and the median item's menu contains 122 listings, a ratio of 1.82×. A long menu contributes more items to this second calculation.

This is a property of the catalogue. It does not show which menus consumers visit most often, because the study contains no browsing or impression records.

Long menus add sections and deepen them

A long menu is more than a short menu with extra choices in the same categories. Across the size bands, both the number of sections and the number of items within a section rose. Section count increased more sharply.

From the smallest band to the largest, median section count grew 7.7× and median section depth grew 3.7×. The overall median of 6 listings per section does not mean every menu size has the same section depth.

For a consumer, this suggests a practical starting point: use the section that matches the meal you intend to order. That recommendation follows the menu structure; the data does not measure how much time it saves.

Sections were counted using their recorded names. Similar names were not merged, and empty category labels could count as a section, so these section counts are an upper bound on consistently defined categories.

The same restaurant can have different menus across apps

Comparing each app's overall menu median would mix different restaurant rosters. To make the comparison more useful, DealMeal matched physical branches appearing on both apps and compared their latest recorded menus.

These figures are medians of branch-by-branch differences. They are not the result of subtracting each app's overall median, and they do not mean the ordering holds at every branch. The observations were also not necessarily recorded on the same day.

More listings do not always mean more dishes

In the separate dish-only paired analysis, DoorDash's advantage over Uber Eats fell to zero dishes at the median, across 597 branches. The result's statistical significance depended on how branches with no dish listings on one app were handled, so the defensible conclusion is that the gap largely disappears when counting dishes.

Grubhub's shorter menus persisted in the dish-only comparisons: a median 13 fewer dishes than DoorDash across 535 branches, and 14 fewer than Uber Eats across 556. These dish-only samples differ from the all-listing samples above.

This is a comparison of availability and listing structure, not a price ranking. To examine the price question separately, see our analysis of menu prices for the same restaurant across apps. For an intended order, check whether the dish and portion are available before comparing promotion terms and the final cart.

Does a bigger menu come with more deals?

Among the 1,482 kitchen stores that appeared in both the menu and promotion datasets, menu length showed no meaningful relationship with the number of distinct offer titles. The Spearman correlation was −0.012, with p = 0.64. The same near-zero pattern held within Uber Eats, DoorDash, and Grubhub separately.

The median matched store had 1 distinct offer title; 898 stores, or 60.6%, had exactly one. A larger catalogue was therefore not a useful signal for finding a greater number of distinct promotions in this sample.

Every matched store already had a promotion. The dataset cannot tell us whether a long-menu restaurant is more likely to run any promotion than a restaurant running none. It also cannot establish how many offers were available simultaneously.

The unit matters. Counting repeated daily observations of offers instead of distinct titles produced a small positive correlation of 0.069, with p = 0.008. That measure gives persistent offers more weight, so it answers a different question. Neither result supports a claim that longer menus cause more deals.

The dates also differ: offers were observed July 25–September 2, 2026, while menu observations cover September 7–October 12. This is a historical comparison of matched stores, rather than a test of menus and promotions visible together at checkout.

Why “select items” deserves a closer look

Menu length and promotion specificity are separate issues. An order-level offer asks you to check the basket conditions. An item-specific offer also asks you to identify the eligible food.

In the broader promotion dataset, 10,655 of 253,204 distinct daily offer records, or 4.21%, used “select items” wording. None of those titles named a specific item. Across all daily offer records, 4,523, or 1.79%, named a menu item, and all were on Grubhub.

None of the 225,003 DoorDash and Uber Eats daily offer records named a menu item in the title. This describes the recorded title text, not everything an app may show after opening a promotion. The title alone may leave eligibility unresolved.

The study does not establish that select-items offers belong to longer menus. An apparent relationship in the combined sample disappeared within the platforms that had enough stores to test. Menu length therefore should not be used to guess whether an offer is item-restricted.

Descriptions and photos are separate information gaps

Across all 473,248 latest-observation listings, including shops, 75.71% had a description and 62.11% had an image. Missing information can make an item harder to evaluate in principle, but those coverage figures do not measure consumer hesitation, scrolling, or conversion.

How to use these findings when choosing an order

1. Start with the food you intend to buy. Use the relevant section and compare the same dish, portion, and branch across apps where possible.

2. Read the promotion scope. Check whether it applies to the order or selected items, then verify that your intended food qualifies.

3. Check the conditions before adding extras. Review minimum spend, discount caps, and any account or ordering-channel restrictions shown with the live offer.

4. Compare the final cart. A longer menu and a larger headline promotion do not establish a lower checkout total. Review the actual food, discount, fees, taxes, and other charges shown.

DealMeal can help you discover and compare available offers. Our guide to using DealMeal explains the process. Verify the current menu and promotion conditions in the ordering app before placing an order.

Methodology and limitations

The primary DealMeal menu export was prepared October 13, 2026, and contained 512,089 item records across 5,176 platform-restaurants. It covered 36 observation dates between September 7 and October 12. Selecting the latest recorded observation per platform-restaurant retained 473,248 listings.

Listings with at least half their items classified as retail were treated as shops. Excluding 250 flagged shops removed 44,531 listings, leaving the published population of 4,926 menus and 428,717 listings. All listing types count toward menu size; dish-only comparisons are identified separately. Percentiles use the nearest-rank method.

The geographically located portion is concentrated in the Washington DC metropolitan area, including the District, Northern Virginia, and suburban Maryland. However, 14.25% of the latest-observation items had no open branch record and could not be geographically located. These results should not be presented as national, state, or city benchmarks. The DealMeal app works across the United States, while DealMeal.co currently showcases selected regional examples on the web.

• Observation completeness: the records are not a simultaneous snapshot. Combining dates can inflate menu size, while the latest observation can be incomplete. At 70 of 1,562 restaurants observed on multiple dates, the final observation contained fewer items than an earlier one. The kitchen median remained 67 under the completeness check.

• Timestamp gaps: 225 records at 6 platform-restaurants had unreadable dates and were omitted by the latest-observation rule.

• Shop-filter limits: at least one convenience store passed the filter. Its 1,536 listings explain the apparent maximum in the retained population; the longest verified restaurant menu contained 1,082. No corrected distribution excluding every missed shop is available.

• Category limits: section names were not standardized. Similar names and 256 empty-category rows can inflate the number of distinct recorded sections.

• Promotion coverage: the secondary dataset contained 253,204 distinct daily USD offer records, representing 51,572 distinct offers at 44,628 stores over 39 observation days from July 25–September 2, 2026. The matched kitchen sample was 1,482 stores, or 3.3% of those stores, and included only stores with offers.

• Unmeasured outcomes: the study does not measure discovery time, consumer activity, menu growth, realized savings, or whether restaurants without observed promotions had any.

For more context on interpreting proprietary findings, read how DealMeal measures food delivery deals.

Frequently asked questions

How many items are on a typical delivery menu?

The median was 67 listings among 4,926 retained menus in DealMeal's DC-area dataset. The middle half contained 37–114 listings. It is a local study result, not a nationwide estimate.

Does menu size mean the number of different dishes?

No. Total menu size includes drinks, combos, and other listing types. The median number of dish listings was 46, measured separately from the total menu median.

Why does the same restaurant have different menus on delivery apps?

The paired analysis found different listing counts at the same physical branches. It did not identify whether merchant choices, platform presentation, observation timing, or other factors caused those differences.

Do larger restaurant menus have more discounts?

Menu size did not predict the number of distinct offer titles among matched stores that already ran promotions. The study cannot answer whether larger menus are more likely to have any promotion.

Does “select items” mean every item in a menu is discounted?

It signals an eligibility restriction. Check the promotion details and the intended item in the ordering app rather than assuming the entire menu qualifies.

Does a shorter menu make finding deals faster?

This study cannot establish that. It measured menu structure and recorded promotion text, without scroll, tap, impression, or timing data.

Choose the meal, then compare the offer

The typical retained menu had 67 listings, and long menus held a disproportionate share of the catalogue. Yet menu size did not predict more distinct promotions among stores running offers. Use menu length as context, then check the intended dish, eligibility, and final cart.

Get DealMeal to compare available food deals and use the ordering app to confirm the current terms for your meal.