Back to the journalhungry but don’t know what to eat

Hungry but Don’t Know What to Eat? 7 Ways to Narrow It Down

Start with a craving, set a budget, and use five more filters when dinner is still a toss-up.

Written by
Arian
Published
Reading time
10 min read
DealMeal Labs cover reading “Hungry but Don’t Know What to Eat?” beside fried chicken, a grain bowl, and pizza.

Hungry but Don’t Know What to Eat? 7 Ways to Narrow It Down

If you’re hungry but don’t know what to eat, start with one thing you want from dinner. Something crispy? A filling meal? A dish under $15? Each answer gives you a way to stop scrolling past options you were unlikely to order.

DealMeal Labs tested seven ways to narrow the choice using 196 delivery feeds from November 15–16, 2026. The median observed feed contained 3,013 dinner dishes from 69 restaurants with menus in the study. Here’s how much each filter cut that list and when to use it.

Hungry but don’t know what to eat? Start with a craving

Choose something crispy, spicy, cheesy, fresh, or filling. Then set a price cap per dish and use restaurant ratings if you still have too many options. In the feeds we studied, a craving, a $15 cap, and a 4.7-star restaurant minimum left a median of 33–59 dishes from 7–11 restaurants, depending on the craving. If you have a firm deadline, check live arrival estimates first.

How much can you narrow down dinner?

  • One craving removed 77–93% of the measured choices, depending on which craving you picked.

  • A $15 menu-price cap kept 46.9% of dishes. A $10 cap kept 19.8%.

  • A 4.7-star minimum kept 25.8% of dishes overall. The share varied by app.

  • In the DishDating sample, 20 of 30 people who saw deck cards liked at least one. Among those 20, the median first like was on the second card.

DealMeal Labs finding

“Spicy” alone left a median of 350 dishes. Adding a $15 cap brought that down to 155. Adding a 4.7-star restaurant minimum left 33 dishes from seven restaurants. That combination left zero matches in 12 of the 196 feeds, so have a fallback if your shortlist comes up empty.

Median dinner choices after combining a craving, a $15 menu-price cap, and a 4.7-star restaurant minimum

Craving

Craving alone

With $15 cap

With $15 cap and 4.7+ rating

Feeds with zero matches after all three

Crispy

581 dishes, 56 restaurants

360 dishes, 52 restaurants

59 dishes, 11 restaurants

11 of 196

Spicy

350 dishes, 52 restaurants

155 dishes, 40 restaurants

33 dishes, 7 restaurants

12 of 196

Cheesy

375 dishes, 46 restaurants

206 dishes, 41 restaurants

35 dishes, 9 restaurants

10 of 196

Fresh

211 dishes, 34 restaurants

124 dishes, 28 restaurants

35 dishes, 9 restaurants

10 of 196

Filling

722 dishes, 61 restaurants

284 dishes, 43 restaurants

39 dishes, 9 restaurants

4 of 196

Spicy chicken bowl in focus with other dinners blurred behind it and symbols for craving price and rating.
A craving, a $15 price cap, and a 4.7-star restaurant minimum left a median of 33–59 dishes, depending on the craving.

1. Name the craving before choosing the dish

Try a word for how you want dinner to taste or feel. “Crispy” gives you room to choose between fried chicken and a crunchy sandwich. “Spicy” can take you toward wings, curry, or a burrito. You can decide on the cuisine later.

Across the 196 feeds, fresh kept a median of 6.8% of dinner dishes, spicy 11.2%, cheesy 12.8%, crispy 19.2%, and filling 23.1%. DealMeal Labs identified cravings using words in menu listings, so some dishes that fit a craving may have been missed. If you can’t name a craving at all, start with what to eat when nothing sounds good.

2. Set a price cap per dish

A $10 cap kept a median of 19.8% of measured dinner choices, or 683 dishes per feed. A $15 cap kept 46.9%, or 1,594 dishes. Raising the cap to $20 kept 75.5%, or 2,432 dishes. Shares and counts are separate medians across feeds.

If budget is firm, screen by listed dish price and then check the total before ordering. Allow room for taxes, fees, tip, and paid extras where applicable. For a close decision, compare the checkout totals of dishes you’d actually want to eat.

3. Decide when you need to eat

If dinner has to arrive by a specific time, check the current delivery estimate before you open the menu. In historical U.S. quotes from July 13, 2025, through July 2, 2026, the median was 36 minutes on DoorDash and 29 minutes on Grubhub. At dinner, from 5 to 9 p.m. Eastern, the medians were 37 and 29 minutes, respectively. These are archived estimates, so use the quote on your order to make tonight’s decision.

For pickup, consider how far you’re willing to travel. In 129 DoorDash and Grubhub feeds, a two-mile radius kept a median of 26.1% of dishes. A one-mile radius kept 5.6% and left 33 feeds with zero matches. Distances were measured from each observation area’s center. The distance data and historical quotes came from separate datasets, so the analysis could not establish whether closer restaurants delivered faster.

4. Pick one cuisine or menu category

In 130 Uber Eats and Grubhub feeds with cuisine tags, sandwiches kept a median of 22.6% of dishes, chicken 17.0%, pizza 14.7%, salads 12.8%, and burgers 12.2%. Where available, those categories led to medians of seven to 12 restaurants.

Thai and Korean tags were each absent from about four in 10 feeds. Where either tag was available, it led to a median of two restaurants. If your first choice is absent, switch back to the craving behind it. A search for something spicy can keep dinner moving when the Thai options are limited. Tags are assigned at the restaurant level, can overlap, and vary by app.

5. Match the portion to your appetite

If you’re very hungry, look at bowls, burritos, curries, or platters. If you just want a bite, try a soup, salad, starter, or small plate and check the menu description for portion clues. A bowl and a cup of soup can both sound good while doing very different jobs.

In the study, the filling choices outside the light-bite category kept a median of 21.7% of dishes. The light-bite rule, which included fresh dishes and menu categories such as appetizers, soups, salads, and tapas, kept 16.9%. These groups were based on menu wording and categories. Serving sizes were unmeasured.

A diner chooses between a cup of soup and a full chicken and rice bowl.
Decide whether you want a light bite or a full meal before comparing dishes.

6. Use restaurant ratings as a tiebreaker

Once you have a craving and a budget, a rating threshold can cut the number of menus you need to open. A minimum restaurant rating of 4.7 kept a median of 25.8% of dishes overall. It kept 21.0% in DoorDash feeds, 23.3% in Grubhub feeds, and 39.5% in Uber Eats feeds. Ratings were available for 95.6% of observed dishes through their restaurant listings.

The score belongs to the restaurant as a whole. If you want one specific dish, use its description and any dish-specific reviews you can find. If your craving and budget leave zero matches after adding 4.7 stars, relax the rating threshold first.

7. Let an eligible deal make the final cut

If two dishes sound equally good, see whether either qualifies for a deal. The median share of dinner dishes with an observable eligibility signal was 2.6% across 133 DoorDash and Uber Eats feeds, with a separate median count of 123 eligible dishes per feed. Signals included an explicit deal match, a struck-through price, or a menu category for offers such as buy one, get one free or a free item with qualifying spend. Grubhub was excluded from this calculation because the required eligibility marker was unavailable.

Check that the offer applies to your item and order, then compare the final total. If meeting a minimum means buying an extra dish you don’t want, price the original meal on its own before you decide. The DealMeal guide shows how to explore and compare offers.

Three dinner plates with a coral paper frame highlighting the center plate to illustrate deal eligibility.
Check which dishes qualify for the offer before comparing checkout totals.

Still stuck? Try browsing one dish at a time

DishDating, DealMeal’s swipe feature, gives you a dish to react to before you see the next one. Of 30 people who saw recorded deck cards from October 19 through November 16, 2026, 20 liked or starred at least one. Among those 20, the median first positive reaction came on the second card, with the middle half spanning the first through seventh cards. The other 10 saw a median of 20 cards without liking or starring one.

Across the broader October 6–November 16 action sample, likes and stars accounted for 33.3% of 759 like, star, or dislike records from 31 people. This was a small early sample. The results describe card reactions and leave open whether people ordered the dish, were happy with dinner, or decided faster than they would have using a menu.

Methodology and limitations

DealMeal Labs paired deal-feed listings observed on November 15–16, 2026, Eastern time, with menus from November 13–16. The study covered Washington, DC, and 13 large metro areas. The 196 feeds each contained at least 200 matched dinner dish listings: 66 DoorDash, 67 Uber Eats, and 63 Grubhub. The menu dataset contained 1,190,044 rows and yielded 511,585 qualifying dinner dish listings at 11,531 restaurant listings. A dish could appear in multiple feeds, producing 651,049 dish observations across the 196 feeds.

Qualifying dishes were priced $5–$60 and excluded drinks, desserts, sides, breakfast, and packaged items. A feed included restaurants showing a deal, plus listings such as grocery and convenience stores. Menus were available for 29.6% of distinct feed listings and 37.4% of feed-by-listing combinations. The measured dish and restaurant counts are lower bounds for those feeds. The DealMeal app works across the United States, while DealMeal.co showcases selected regional examples on the web.

Filter shares are medians of the shares measured feed by feed. Combined filters were applied directly to the same dishes. Craving categories used menu-word rules with incomplete coverage, and cuisine tags were standardized and counted once per restaurant per category. The archived delivery analysis used the first U.S. reading per restaurant per UTC day, with numeric quotes for 37,396 DoorDash and 29,527 Grubhub restaurant-days.

The DishDating analysis excluded team activity and covered 1,767 action records representing 1,835 events from 48 users. Deck-card analysis excluded photo-grid cards and ordered the remaining cards by time and then deck position. The study measured available choices and recorded reactions. It did not test which filters people use or whether filtering improves satisfaction.

Frequently asked questions

What should I eat when I’m hungry but nothing sounds good?

Start with portion size. Pick a full meal or a light bite, then choose a temperature or texture you’d welcome, such as something warm or crunchy. Compare a couple of dishes that fit.

Is $15 a good dinner budget for delivery?

It can be a practical menu-price cutoff. It kept 46.9% of dishes in the median measured feed. For a $15 total budget, check the cart and leave room for any fees, taxes, or tip that apply.

Which filter should I remove if there are zero results?

Loosen your least important filter. If price is firm, try a lower restaurant rating or a different cuisine. If your time limit is firm, keep the arrival estimate and change the type of dish.

Does a restaurant deal apply to every dish?

Check the offer terms for eligible dishes, minimum spend, and any pickup or delivery requirements. Then confirm the discount appears in your cart.

From shortlist to dinner

Pick a craving, cap the price, and open a few menus. If timing matters, check arrival estimates early. When two dishes both fit, choose by portion and checkout total. If you’d rather react to dishes as they come up, get DealMeal and try DishDating.