Yellowpages-UAE Scraper — UAE Business Directory Leads

Exports any trade category from yellowpages-uae.com as B2B lead rows: business name, city, street address, PO Box, landline, mobile, e-mail, brands and branch count.

This is not yello.ae. The other UAE directory scrapers on this store all target yello.ae, a different site. This one reads yellowpages-uae.com, and as far as the store search shows, nothing else does.

---

Coverage

Measured on a live run, 2026-09-13: the auto spare parts category declares 1,254 businesses and this Actor delivered 1,252 of them — 99.8% — across 63 pages, stopping cleanly at the end of the category rather than at an arbitrary page limit.

Category sizes vary a lot. auto-spare-parts runs to 63 pages; car-rental ends at page 39. The Actor pages until the directory says there is nothing more, and reports what it got against what the category declared, per category, in RUN_SUMMARY.

The e-mail field, honestly

Roughly half the e-mail addresses in this directory are the directory's own. Listings that have not bought a contact upgrade show yellowpagesfreelistings@… instead of the business's address. Measured across two categories at depth: 54.8% and 63.3% of non-empty e-mails.

This Actor removes them. You get email: null and emailWasDirectoryInbox: true, so you can see why the field is empty. Ship them and a buyer mail-merges hundreds of messages into the directory's own inbox.

After removal, a genuine business e-mail is on about 40% of listings.

Measured fill — stratified by depth, not read off page one

Page 1 of every category is the paid advertising tier and is close to 100% on everything. It is not representative, and a fill table built from it is wrong. These figures come from pages 1, 5, 20, 40 and 63 across two categories (n=153):

field	overall	page 1 (ads)	deeper pages
name · city	100%	100%	100%
phone (landline)	88–91%	68–70%	97%
streetAddress	88–95%	100%	85–95%
poBox	72%	36–58%	77–90%
email (after removing the house address)	~40%	high	low
mobile	32–42%	75–100%	10–27%
latitude / longitude	37%	50–74%	27–33%
phone2	3–20%	10–100%	0%

The landline is the field to build on. It is the only contact that holds up at depth, at ~97%. If you need e-mail specifically, turn on requireEmail and you will be charged only for rows that carry a real one.

Every row carries hasEmail, hasMobile, hasPhone, hasGeo and a contactTier (email / mobile / landline / none), so a null is never ambiguous.

Input

input	default	what it does
categorySlugs	["auto-spare-parts"]	Slugs from the category URL, …/uae/<slug>
emirate	—	Optional city filter (Dubai, Sharjah, Abu Dhabi, Ajman, RAK, UAQ, Fujairah)
requireEmail	false	Only rows with a real e-mail
requireAnyContact	false	Only rows with an e-mail, mobile or landline
maxRecords	0 (all)	Cap on delivered and charged rows
proxyConfiguration	Apify Proxy	Datacenter is fine

One row per business, deduplicated on the directory's own branch id — so a business appearing in two categories you requested is delivered once, not twice.

Why a run either completes or fails

The site issues a short-lived session cookie (ip_verified, 30 minutes). Without it a crawl is cut off after roughly 15 requests with a site-wide 403. The Actor mints and refreshes that cookie itself.

That matters for what you are charged. A 403 page contains no business records — so a scraper that stops "when a page returns no new records" would end a third of the way through a category, report success, and bill you for it. This Actor treats any non-200 as an error, not as the end of the data: it re-mints the session and retries, and if it still cannot read the category it fails the run and delivers nothing rather than selling you part of a directory.

It also checks its own parsing: the page's record count is compared against what was decoded, and a mismatch fails the run.

Pricing

Pay per event: $0.0004 per business-lead — charged only for a row delivered to your dataset. Filters are applied before delivery, so rows they drop are never charged, and a run that returns nothing costs nothing. The complete auto-spare-parts category, 1,252 businesses, is $0.50.

The whole category has to be paged before it can be filtered, so a small capped run costs us about the same as a full one. That is why the default is the whole category — it is the setting where you get the most for your money, not the least.

Source and legal

yellowpages-uae.com is a public business directory; listings are published by the businesses themselves to be found. robots.txt is a named bad-bot list (Baiduspider, PetalBot, YandexBot, ia_archiver, Sogou, Exabot, MojeekBot, DotBot, MJ12bot, BLEXBot and scrapy, the last of which also carries Disallow: /*?*), with no blanket restriction on other agents.

The site's terms state, verbatim:

"You may only use or reproduce the information in the Content for your own personal, non-commercial or educational use."

Copyright is held by Lunad Digital Portal. That is a reuse-licence restriction rather than an access control, and it is your responsibility to decide what your intended use is before you run this. For outreach into the UAE, the Consumer Protection Law's e-marketing rules apply, and the equivalent wherever you are sending from.