Extract Talabat Cuisines and Ratings for MENA Market Research (2026)

Published April 27, 2026 · 1536 words · For researchers

Thirdwatch's Talabat Scraper returns restaurant cuisine tags, ratings, total_ratings, and GPS coordinates across all nine MENA markets at $0.002 per restaurant — UAE, Saudi Arabia, Kuwait, Bahrain, Oman, Qatar, Jordan, Egypt, and Iraq. Built for F&B market researchers, cloud-kitchen operators, and franchise scouts who need structured cuisine and quality data without manual menu reviews or paid subscriptions.

Why use Talabat for MENA F&B market research

The MENA food-delivery market crossed $20 billion in gross merchandise value in 2024, according to Delivery Hero's 2024 annual report, and Talabat is the regional leader by a meaningful margin. For an F&B market researcher, that share matters because it makes Talabat's catalog the most consistent cross-country dataset of delivery-enabled restaurants — the universe that matters when assessing any digital food opportunity from cloud kitchens to franchise expansion.

The job-to-be-done is structured. A cloud-kitchen operator scouting a new market needs to know: how dense is restaurant supply by neighbourhood, what cuisines are over- or under-represented, where are the highly-rated chains and where are the gaps. A franchise scout asks the same questions in different framing: which countries have room for a new pizza chain, where is the nearest Korean BBQ in Riyadh, what's the rating distribution for sushi in UAE. All of these reduce to filtering and aggregating the same Talabat dataset. The actor returns it in structured JSON ready for pandas or BigQuery.

How does this compare to the alternatives?

Three options for MENA F&B market research data:

Industry reports give you context once a year; Talabat gives you the actual restaurant supply on demand. The Talabat Scraper actor page documents every output field, but for market research the highest-value combo is cuisine × latitude/longitude × rating × total_ratings — those four columns answer most strategic questions about where to enter, where to expand, and where to avoid.

How to do MENA F&B market research with Talabat in 4 steps

Step 1: How do I authenticate against Apify?

Sign in at apify.com (free tier, no credit card), open Settings → Integrations, and copy your personal API token. Every example below assumes the token is in APIFY_TOKEN:

export APIFY_TOKEN="apify_api_xxxxxxxxxxxxxxxx"

Step 2: How do I pull the full restaurant catalog for one country?

Use a broad-keyword sweep — generic cuisine terms — to surface the country's restaurant universe. scrapeMenu: false keeps the run fast since you only need restaurant-level metadata for cuisine analysis.

import os, requests, json, pathlib

ACTOR = "thirdwatch~talabat-scraper"
TOKEN = os.environ["APIFY_TOKEN"]

SWEEP = ["pizza", "burger", "sushi", "biryani", "shawarma", "kebab",
         "noodles", "salad", "coffee", "dessert", "bakery", "breakfast"]

resp = requests.post(
    f"https://api.apify.com/v2/acts/{ACTOR}/run-sync-get-dataset-items",
    params={"token": TOKEN},
    json={
        "queries": SWEEP,
        "country": "uae",
        "maxResults": 200,
        "scrapeMenu": False,
    },
    timeout=900,
)
restaurants = resp.json()
pathlib.Path("uae-restaurants.json").write_text(json.dumps(restaurants))
print(f"UAE: {len({r['slug'] for r in restaurants})} unique restaurants")

12 queries × 200 results × dedupe by slug typically returns 1,500-2,500 unique restaurants per country.

Step 3: How do I aggregate cuisine distribution and rating density?

Talabat returns cuisine as a comma-separated string ("Pizza, Italian, Fast Food"); explode that into individual cuisine tags first.

import pandas as pd

df = pd.json_normalize(restaurants)
df = df.drop_duplicates(subset=["slug"])
df["cuisine_list"] = df["cuisine"].fillna("").str.split(", ")
exploded = df.explode("cuisine_list")
exploded = exploded[exploded.cuisine_list.str.strip() != ""]

# Cuisine share + average rating (only for restaurants with 100+ ratings)
quality = (
    exploded[exploded.total_ratings >= 100]
    .groupby("cuisine_list")
    .agg(count=("slug", "nunique"),
         avg_rating=("rating", "mean"),
         med_total_ratings=("total_ratings", "median"))
    .sort_values("count", ascending=False)
)
print(quality.head(20))

The output is a clean cuisine leaderboard. Cuisines with low count and low avg_rating are gap candidates; cuisines with high count and high avg_rating are saturated.

Step 4: How do I cluster restaurants by GPS to find delivery hot zones?

DBSCAN clustering at ~500m radius reveals natural delivery hubs, with each cluster's cuisine mix and average rating telling you the area's character.

from sklearn.cluster import DBSCAN
import numpy as np

geo = df.dropna(subset=["latitude", "longitude"]).copy()
coords = np.radians(geo[["latitude", "longitude"]].values)
db = DBSCAN(eps=0.5/6371, min_samples=10, metric="haversine").fit(coords)
geo["cluster"] = db.labels_

cluster_summary = (
    geo[geo.cluster >= 0]
    .groupby("cluster")
    .agg(
        n_restaurants=("slug", "nunique"),
        avg_rating=("rating", "mean"),
        center_lat=("latitude", "mean"),
        center_lng=("longitude", "mean"),
    )
    .sort_values("n_restaurants", ascending=False)
)
print(cluster_summary.head(15))

Clusters are natural delivery hubs — Dubai Marina, Downtown, JBR. Areas between clusters at population centres are the dark-kitchen opportunity.

Sample output

A single restaurant record (UAE, menu disabled for speed) looks like this. 1,500 such rows weigh under 1 MB and ingest into pandas in milliseconds.

{
  "name": "Pizza Hut",
  "slug": "pizza-hut",
  "cuisine": "Pizza, Italian, Fast Food",
  "rating": 4.2,
  "total_ratings": 15420,
  "latitude": 25.2048,
  "longitude": 55.2708,
  "accepts_cash": true,
  "accepts_card": true,
  "most_selling_items": ["Pepperoni Pizza", "Margherita", "Chicken Supreme"],
  "country": "uae",
  "url": "https://www.talabat.com/uae/pizza-hut",
  "scraped_at": "2026-04-27T08:14:22.000000+00:00"
}

cuisine is the analysis backbone; total_ratings is the trust filter. most_selling_items is an underrated field — three popular dishes per restaurant aggregated across 1,500 restaurants in a country reveals what people in that market actually buy, which often differs from what the cuisine tag implies.

Common pitfalls

Three things go wrong in F&B research pipelines on Talabat data. Cuisine-tag overlap — a restaurant tagged "Pizza, Italian, Fast Food" appears in three cuisines after explode; this is correct for share-of-supply analysis but inflates raw counts. Rating sample-size confusion — a brand-new entrant with 5 ratings averaging 4.9 is not better than a 4.2-rated chain with 15,000 ratings. Always filter on total_ratings >= 100. GPS clustering eps tuning — 500m is right for dense urban areas like Dubai or Riyadh; for less dense markets like Oman or rural Egypt, expand to 1.5-2km or you'll get hundreds of single-restaurant "clusters".

Thirdwatch's actor returns total_ratings and latitude/longitude on every record so the right filtering and clustering can happen downstream. The pure-HTTP architecture means you can sweep all nine countries in a single afternoon for under $20. A fourth pitfall to flag: cuisine tags are written by the restaurant operator and are inconsistent across stores of the same chain — McDonald's UAE may tag "Burgers, American, Fast Food" while McDonald's Egypt tags "American, Burgers, Sandwiches"; canonicalise to a tag-set before comparing across countries.

Related use cases

• Scrape Talabat restaurant menus for price monitoring
• Build a food delivery price comparison with Talabat
• Monitor Talabat promotions and discounts at scale
• The complete guide to scraping food delivery platforms
• All Thirdwatch use-case guides

Frequently asked questions

Why is Talabat the right dataset for MENA F&B market research?

Talabat is the dominant food-delivery aggregator in nine MENA countries, with the most consistent restaurant catalog in the region. Other aggregators have spotty coverage by country. A single dataset spanning UAE through Iraq, with cuisine tags, ratings, and GPS coordinates, is the cleanest input to any cross-country F&B market analysis.

What does cuisine distribution tell me about market opportunity?

Under-represented cuisines relative to population demand are the cleanest gap signal. If Vietnamese restaurants are 0.5% of UAE listings versus 3% in similar global markets, the gap is real opportunity for a cloud kitchen or franchise. Combine with rating distribution to see whether the existing players are weak — low average rating in a niche cuisine is the strongest enter-this-market signal.

How accurate are Talabat ratings as a quality proxy?

Talabat ratings are aggregated customer scores with sample sizes that vary widely — 4.2 with 15,000 ratings is meaningful, 4.8 with 12 ratings is not. Always filter on total_ratings before reading rating; below 100 ratings the value is noise. The actor returns both fields so you can apply the filter downstream.

Can I cluster restaurants by GPS to identify food-delivery hot zones?

Yes. Every restaurant record includes latitude and longitude. Cluster with DBSCAN at ~500m radius to identify natural delivery hubs, then count per-cluster restaurant density and cuisine diversity. The result is a heatmap of where existing F&B supply is concentrated — and the gaps between clusters are where new dark-kitchen real estate makes sense.

How is Talabat data different from Google Maps for the same purpose?

Google Maps gives broader business coverage including non-delivering venues, but its cuisine tagging is inconsistent and ratings often blend tourist visits with regular customers. Talabat is delivery-only but tags cuisine consistently, surfaces popular items per restaurant, and ratings reflect repeat-customer behaviour. For F&B market research, Talabat is more focused; for general business density, Google Maps is broader.

Does Talabat data cover the full restaurant universe in each country?

Talabat's MENA share is dominant but not absolute — small or independent restaurants that don't enable delivery may be missing. Treat Talabat as the delivery-enabled supply view, which is the relevant universe for any digital F&B opportunity. For total restaurant counts cross-reference with Google Maps or local government databases.

Run the Talabat Scraper on Apify Store — pay-per-restaurant, free to try, no credit card to test.