Salla & Zid App Store Tracker

Every app in the Salla App Store (apps.salla.sa) and the Zid App Market (apps.zid.sa) — the two app stores behind most online shops in Saudi Arabia and the Gulf — with the developer company, category, rating, number of ratings and price.

Turn on Only new or changed apps, schedule it, and each run returns only the apps that were just listed or whose rating or price moved.

No API key, no login, no browser, no proxy.

---

What you get

One row per app, 22 fields, every one always present. Missing values come back as null, never dropped. Example values from a run on 15 Sep 2026:

Field	Example
platform / appId	salla / 939200655
name	Noti Money
developer	وادي بلسم لتقنية المعلومات
category / categories	Marketing / ["Marketing", "Analytics", "Chat", "Others"]
rating / ratingCount	4.8 / 297
pricingText	69 SAR / Monthly (Free Trial 7 Days)
price / currency / isFree	69.0 / SAR / false
url	https://apps.salla.sa/en/app/939200655
changeType / changedFields	UPDATED / ["rating", "ratingCount"] — only with Only new or changed apps
previousRating / previousRatingCount / previousPrice	the values at the last run — only with UPDATED

Plus shortDescription, iconUrl, lastUpdated and scrapedAt.

developer is a business, never a person. Many developers list only their own name. When a listing shows a person's name instead of a company, developer is null. This is decided by a rule, not a registry: a name with no business word (company, LLC, Est., شركة, مؤسسة, …) that starts with a common given name is treated as a person. On 15 Sep 2026 it hid 74 of 802 Salla apps' developers and 35 of 326 Zid apps' developers. No developer email, phone or reviewer name is collected.

rating is null for apps nobody has rated. Both stores display a default score for unrated apps (Zid shows 5 stars); that is not a rating.

price is the amount in the pricing text, 0 for free apps, and null when the store shows no amount — free trials and usage-based plans.

---

Input

Input	Default	What it does
platforms	["salla", "zid"]	Which stores to read.
language	en	en or ar — names, categories and pricing text as each store shows them in that language.
maxAppsPerPlatform	5000	Stop after this many apps from each store, in the store's own order. With onlyChanges it counts apps checked.
onlyChanges	false	Tracking mode — see below.
stateStoreName	gulf-app-stores-state	Named key-value store that keeps the snapshot between runs. One name per independent tracker.

---

Tracking: only new or changed apps

Turn on onlyChanges and schedule the Actor. Each run compares every app with the last run and delivers only:

• NEW — an app not seen before (the first run delivers every app as NEW), or
• UPDATED — its rating, ratingCount, price or pricingText changed. changedFields says which; previousRating, previousRatingCount and previousPrice hold the values from the last run.

A quiet day costs nothing in results: measured 15 Sep 2026, the second run over both stores delivered 0 rows in 90.3 s.

Honest limits of tracking:

• An app removed from a store is not reported; it simply stops appearing.
• The snapshot is kept per store and per language. Switching language on the same stateStoreName starts a fresh snapshot.
• An app cut off by your spending limit is left out of the snapshot, so it comes back as NEW next run rather than being missed.

---

Use it for

• Market research on the Saudi and Gulf e-commerce app ecosystem — which categories are crowded, what apps charge, who has traction.
• Competitor tracking for app developers — rating and review growth, price changes, new entrants in your category.
• Partnership and sales lists — the companies building for Salla and Zid merchants.

---

How it works

Both store websites load their app lists from a public JSON API in the visitor's browser. The Actor reads the same API, over plain HTTP:

1. Salla: the app list (15 apps a page), then each category's app list, so every app gets the categories it appears under.
2. Zid: the app list, 100 apps a page. Zid's list repeats some apps; each app is delivered once.

Measured 15 Sep 2026 on a home connection (speed on Apify's servers will differ):

Run	Apps delivered	Requests	Time
Both stores, first run	1,128 (802 Salla + 326 Zid)	169	94.1 s
Both stores, onlyChanges, second run, nothing changed	0	169	90.3 s
Both stores in Arabic, 5 apps each	10	112	57.9 s

Zero failed requests in all three.

Why a full run takes about a minute and a half. Salla's API allows 240 requests a minute (its x-ratelimit-limit header; the count resets every 60 s) and answered HTTP 429 to a run started right after two unpaced runs. The Actor now sends at most 120 requests a minute, and if a store still answers 429 it waits (honouring Retry-After) instead of giving up. Every Salla run walks the category lists in full, so a small maxAppsPerPlatform still costs about 110 requests.

---

Honest notes

Not from the websites' HTML. apps.salla.sa's sitemap names only 29 apps and its pages hold only a few, so the Actor uses the JSON the site itself loads. These APIs are not documented for public use and may change without notice. When a store's list cannot be read to the end, the run report says so (listingComplete: false) and the status message names the store.

robots.txt is respected on both API hosts (checked 15 Sep 2026: both allow general crawling).

Categories on Salla: category is the first category other than "Others"; categories lists all. 4 of 802 apps were in no category list and have category: null.

Ratings are rounded by the stores' lists: Salla to one decimal, Zid to a whole star.

Currency on Zid follows the location the request comes from, and the store sends no currency choice of its own. A run from Kuwait saw KWD; a run on Apify's servers (15 Sep 2026) got amounts with no currency mark, such as From 126.50 / 1 Month, so currency is null there and the amount's currency is unknown. Salla prices always carry SAR. Only 54 of 326 Zid apps show an amount at all; the rest are free, trials or usage-based.

lastUpdated is always null today. Neither store publishes an update date in its listing or app details (checked 15 Sep 2026). The field is kept so the schema does not change if one appears.

Every run publishes a report to the key-value store under RUN_REPORT: apps listed and delivered per store, new/updated/unchanged counts, failed requests, whether each list was read to the end, and a summary — free apps, rated apps, median rating, apps without a category, developers hidden as personal names.

---

Questions, or an app that is missing?

Open an issue with the app's store link.