DEV Community: Mena489

How to Analyse Competitor Website Traffic (Without a $200/Month Tool)

Mena489 — Mon, 24 Aug 2026 20:01:35 +0000

Knowing how much traffic a competitor gets, where it comes from, and which keywords drive it is the foundation of any serious SEO or market-entry decision. The tools that provide it — Similarweb, Semrush, Ahrefs — start around $130–$500 per month. This guide shows how to get the same core metrics for $10.

Short answer: feed domains to the Website Traffic Analysis tool and get global rank, monthly visits, engagement metrics, traffic-source breakdown and top keywords per domain.

What the data actually tells you

Traffic estimates are estimates. No third-party tool sees a competitor's analytics. What they do see is panel data, clickstream samples and public signals, modelled into a number. Treat the absolute figure as approximate and the relative figures as reliable — if the tool says competitor A gets 3× competitor B, that ratio holds up far better than either raw number.

With that caveat, four metrics carry most of the decision-making weight:

Traffic sources. A site that is 80% paid search is a fundamentally different business from one that is 80% organic. The first stops the day the ad budget stops.
Engagement. Bounce rate and pages-per-visit tell you whether that traffic is qualified or bought cheap.
Top keywords. The terms actually sending them traffic — your keyword research, already done.
Trend direction. Three months of visit history shows whether they are growing or sliding.

Step 1 — Give it domains

{
    "urls": [
        "https://docs.apify.com/",
        "https://competitor-one.com",
        "https://competitor-two.com"
    ]
}

That is the whole input. Pass one domain or a list of fifty.

Step 2 — Read the results

{
    "SiteName": "docs.apify.com",
    "Title": "Apify Documentation",
    "Category": "computers_electronics_and_technology/programming_and_developer_software",
    "GlobalRank": { "Rank": 68321 },
    "CountryRank": { "Country": 840, "CountryCode": "US", "Rank": 41902 },
    "EstimatedMonthlyVisits": {
        "2026-04-01": 812340,
        "2026-05-01": 798120,
        "2026-06-01": 803455
    },
    "Engagments": {
        "Visits": "803455",
        "TimeOnSite": "214.7",
        "PagePerVisit": "3.42",
        "BounceRate": "0.41"
    },
    "TrafficSources": {
        "Direct": 0.41,
        "Search": 0.44,
        "Social": 0.03,
        "Referrals": 0.11,
        "Mail": 0.01
    },
    "TopCountryShares": [ { "Country": 840, "CountryCode": "US", "Value": 0.22 } ]
}

Two things to note: Engagments is spelled that way in the output (an upstream quirk — match it exactly in your code), and TimeOnSite is in seconds, so 214.7 means about three and a half minutes.

Reading it like an analyst

Compute traffic per source, not just share. A 44% search share on 800k visits is 352k organic sessions. Against a competitor with a 70% search share on 90k visits — 63k sessions — you are looking at a site with a smaller organic footprint despite the higher percentage. Percentages mislead; multiply them out.

for site in results:
    visits = float(site["Engagments"]["Visits"])
    for source, share in site["TrafficSources"].items():
        print(f'{site["SiteName"]:30} {source:10} {int(visits * share):>10,}')

Watch the trend, not the snapshot. EstimatedMonthlyVisits gives you three months. A domain sliding 5% a month is losing to someone — worth knowing who before you copy their playbook.

Cross-reference bounce rate with source. High paid share plus high bounce rate is a site buying traffic that does not convert. That is an opening, not a threat.

Calling it from your own code

Python

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("mina_safwat/website-traffic-analysis").call(
    run_input={"urls": ["https://competitor-one.com", "https://competitor-two.com"]}
)

for site in client.dataset(run["defaultDatasetId"]).iterate_items():
    visits = float(site["Engagments"]["Visits"])
    organic = visits * site["TrafficSources"]["Search"]
    print(f'{site["SiteName"]}: {organic:,.0f} organic visits/mo')

cURL

curl -X POST "https://api.apify.com/v2/acts/mina_safwat~website-traffic-analysis/run-sync-get-dataset-items?token=<YOUR_APIFY_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"urls":["https://competitor-one.com"]}'

What it costs

$10 per month, flat, with unlimited domains. Not per lookup, not per credit — a flat monthly rental. Analysing 500 competitor domains costs the same as analysing one.

That is where the economics differ from the incumbents: Similarweb and Semrush price per seat and gate the interesting data behind higher tiers. If all you need is the traffic and keyword layer, you are paying for a whole suite to use one panel of it.

Common use cases

Market entry. Size the players in a category before committing to it.
Competitive SEO. Pull rivals' top keywords and find the terms you are absent from.
Investment due diligence. Sanity-check the traffic claims in a pitch deck.
Partner and acquisition screening. Verify a prospective partner's reach is real.
Client reporting. Drop competitor benchmarks into a monthly deck automatically.

FAQ

How accurate are the traffic estimates?
They are modelled, not measured, and are most reliable for sites above roughly 50k monthly visits. Below that, sample sizes get thin and error bars widen considerably. Use them comparatively.

Does it work for any domain?
Any public site. Very small or brand-new domains may return sparse data simply because there is not enough signal to model.

Can I schedule it?
Yes — schedule the run in Apify Console and have it write to the same dataset each month for a rolling trend of your whole competitive set.

Does it give backlinks?
No. This is traffic, engagement, sources and keywords. Backlink graphs are a different product.

Is this legal?
Yes. You are reading publicly published traffic estimates about public websites — no personal data, no login, no protected content.

Try it: Website Traffic Analysis on Apify Store — $10/month, unlimited domains.

How to Scrape Google Maps for Local Business Leads

Mena489 — Mon, 24 Aug 2026 20:00:20 +0000

Google Maps is the largest verified database of local businesses in existence — names, addresses, phone numbers, websites and ratings for practically every company with a storefront. This guide shows how to extract it into a CSV you can actually work from, in about a minute.

Short answer: give the Google Maps Scraper a search query like marketing agencies near New York and it returns every matching place with contact details.

Why not just use the Google Places API?

You can, and for some jobs you should. But for lead generation it has three hard edges:

1. It charges per request, and the useful fields cost extra. Phone number and website live in a higher-priced tier. A few thousand lookups gets expensive quickly.

2. It caps results per query. You get 20 per page, 60 total, with mandatory delays between pages. There is no way to pull "every dentist in Chicago" in one call.

3. It needs billing setup, a project, and API key management before you can evaluate whether the data is even useful to you.

Scraping the Maps front end sidesteps all three. The trade-off is that you are responsible for handling the page structure, which is what the scraper does for you.

Step 1 — Describe what you want in plain language

{
    "queries": [
        "restaurants in Miami",
        "marketing agencies near New York",
        "dentists in Chicago"
    ],
    "language": "en",
    "country": "us",
    "maxResultsPerQuery": 0,
    "maxConcurrency": 3
}

maxResultsPerQuery: 0 means no limit — pull everything Maps will show for that query. Set a number when you are testing and do not want to pay for a full sweep.

Queries are just what you would type into Maps. "plumbers in Austin TX", "gyms near 90210", and "coffee shops Shoreditch London" all work.

Step 2 — Read the results

{
    "id": "ChIJN1t_tDeuEmsRUsoyG83frY4",
    "title": "Joe's Stone Crab",
    "category": "Seafood restaurant",
    "address": "11 Washington Ave, Miami Beach, FL 33139",
    "phoneNumber": "(305) 673-0365",
    "completePhoneNumber": "+1 305-673-0365",
    "domain": "joesstonecrab.com",
    "url": "https://www.joesstonecrab.com/",
    "coor": "25.768,-80.135",
    "stars": 4.4,
    "reviews": 6128,
    "source_query": "restaurants in Miami"
}

Two fields worth knowing about:

completePhoneNumber is E.164-formatted and is what you want for any dialler, CRM import or WhatsApp workflow. phoneNumber is the display string Maps shows.
source_query tells you which of your queries produced the row. When you run 50 queries at once, this is how you attribute results back to a segment.

Turning results into a qualified lead list

Raw Maps output is not a lead list yet. Two filters do most of the work:

Businesses with no website. If domain is empty, that business has a Maps listing but no site — the single highest-converting segment for web design and digital marketing agencies.

no_website = [p for p in places if not p.get("domain")]

Businesses with low review counts. Under ~20 reviews usually means nobody is managing their online presence, which is the opening for reputation and SEO services.

underserved = [p for p in places if (p.get("reviews") or 0) < 20 and p.get("stars", 0) >= 3.5]

From there, feed domain into the Contact Details Scraper to pull emails and social profiles off each site, and you have a complete outbound list.

Calling it from your own code

Python

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("mina_safwat/google-maps-scraper").call(
    run_input={"queries": ["marketing agencies near New York"], "country": "us"}
)

places = list(client.dataset(run["defaultDatasetId"]).iterate_items())
prospects = [p for p in places if not p.get("domain")]
print(f"{len(prospects)} of {len(places)} have no website")

cURL

curl -X POST "https://api.apify.com/v2/acts/mina_safwat~google-maps-scraper/run-sync-get-dataset-items?token=<YOUR_APIFY_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"queries":["dentists in Chicago"],"country":"us"}'

What it costs

Per place returned: $0.003 on the free plan, down to $0.001 on Gold and above. A thousand local businesses costs $1–$3 depending on your plan.

For comparison, buying a comparable local business list from a data broker runs $0.10–$0.50 per record, and the data is usually months stale. Maps data is as current as the business's own listing.

Common use cases

Agency prospecting. Find businesses in your service area with no website or a neglected one.
Franchise and site selection. Map competitor density by category before signing a lease.
Market sizing. Count how many businesses of a type operate in a region.
CRM enrichment. Match existing accounts against Maps to fill in missing phone numbers and current addresses.

FAQ

How many results can I get per query?
Set maxResultsPerQuery: 0 for no cap. Google itself limits how deep a single query scrolls, so for exhaustive coverage of a large city, split into neighbourhood-level queries rather than one city-wide one.

Does it return reviews?
It returns the star rating and review count. Review text is a separate job.

Is scraping Google Maps legal?
Business names, addresses and phone numbers are factual public information, and scraping publicly accessible data is generally permitted in many jurisdictions. That is not legal advice. Google's Terms of Service prohibit automated access, and if you contact the businesses you collect, you are subject to CAN-SPAM, GDPR, and local telemarketing rules. Check your obligations before you start emailing or calling.

Can I get results outside the US?
Yes. Set country to any two-letter code and language to match — {"country": "de", "language": "de"} for Germany.

Why are some phone numbers missing?
Not every listing publishes one. Filter on completePhoneNumber when a dialler workflow requires it.

Try it: Google Maps Scraper on Apify Store — pay per place, from $0.001.

How to Scrape LinkedIn Profiles Without Cookies or a Session Token

Mena489 — Mon, 24 Aug 2026 19:58:59 +0000

You can extract public LinkedIn profile data — name, headline, location, follower count, full work history, education and recent posts — without a li_at cookie, a logged-in account, or a headless browser. This guide covers why cookie-based LinkedIn scraping keeps breaking, and the approach that does not.

Short answer: send profile URLs to the LinkedIn Profile Scraper and get structured JSON back. No cookie required.

Why cookie-based LinkedIn scraping fails

Almost every LinkedIn scraper on the market asks you for your li_at session cookie. That design has three problems, and all of them eventually cost you the account.

1. The cookie is your account. Handing li_at to a scraper is handing over full session access. If the tool leaks it, someone else is logged in as you.

2. LinkedIn rate-limits per account, not per IP. Rotating proxies does not help when the identity is constant. LinkedIn tracks request volume against the session, and an account pulling a few thousand profiles gets restricted — often permanently, and often on a real account someone uses professionally.

3. Cookies expire, and they expire silently. A pipeline built on li_at works until it does not, usually mid-run, usually returning HTTP 999 or an empty page rather than a clean auth error.

The result is a scraper that needs a human to babysit it, plus a steady supply of burner accounts.

The approach that avoids the account entirely

LinkedIn serves a genuine public renderer for profiles that are visible to logged-out visitors — that is how profiles show up in Google results. The trick is convincing LinkedIn you are an ordinary logged-out reader rather than a bot.

That means getting several things right at once:

A coherent browser identity. TLS fingerprint, User-Agent and sec-ch-ua headers must all describe the same browser. Mismatched values are the single most common tell.
Plausible provenance. A request claiming same-origin with zero LinkedIn cookies is a contradiction. Arriving from a search engine is how a public profile actually gets read.
Country rotation. Retrying from a different country spreads load across IP pools rather than hammering one.

The scraper handles all of it. You send URLs; it deals with the rest.

Step 1 — Give it profile URLs

{
    "profile_urls": [
        "https://www.linkedin.com/in/williamhgates/",
        "https://www.linkedin.com/in/satyanadella/"
    ],
    "proxy_country": "US,GB,DE,CA,FR",
    "max_concurrency": 5
}

proxy_country accepts a comma-separated list. Rotating across countries per retry measurably reduces 999 and 404 rates compared with pinning every attempt to one country.

Step 2 — Read the results

{
    "url": "https://www.linkedin.com/in/williamhgates/",
    "status": 200,
    "name": "Bill Gates",
    "headline": "Chair, Gates Foundation and Founder, Breakthrough Energy",
    "location": "Seattle, Washington, United States",
    "country": "US",
    "follower_count": 40304014,
    "experience": [
        { "name": "Gates Foundation", "url": "...", "start_year": 2000, "end_year": null }
    ],
    "education": [
        { "name": "Harvard University", "url": "...", "start_year": 1973, "end_year": 1975 }
    ],
    "recent_posts": [
        { "text": "...", "url": "...", "published_at": "...", "likes": 1203 }
    ]
}

An end_year of null means the role is current — useful for filtering to people actively at a target company.

Calling it from your own code

Python

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("mina_safwat/linkedin-profile-scraper-no-cookies").call(
    run_input={
        "profile_urls": ["https://www.linkedin.com/in/williamhgates/"],
        "proxy_country": "US,GB,DE,CA,FR",
    }
)

for profile in client.dataset(run["defaultDatasetId"]).iterate_items():
    current = [job for job in profile.get("experience", []) if job.get("end_year") is None]
    print(profile["name"], "→", current[0]["name"] if current else "no current role")

cURL

curl -X POST "https://api.apify.com/v2/acts/mina_safwat~linkedin-profile-scraper-no-cookies/run-sync-get-dataset-items?token=<YOUR_APIFY_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"profile_urls":["https://www.linkedin.com/in/williamhgates/"]}'

What it costs

Per profile returned: $0.0075 on the free plan, $0.00583 on Bronze, $0.00417 on Silver, and $0.0025 on Gold and above. Enriching a 5,000-row prospect list costs about $12.50 on Gold.

Compare that with LinkedIn Sales Navigator at roughly $99/month per seat, which still does not let you export data programmatically.

Common use cases

B2B lead enrichment. Turn a list of profile URLs into titles, companies and tenure for scoring and routing.
Recruitment sourcing. Filter candidates by current employer, years in role, or education without paying per InMail.
Sales prospecting. Detect job changes — a new VP in your ICP is the strongest buying signal there is.
Market mapping. Chart who works where across an industry segment.

FAQ

Do I really not need a cookie?
Correct, for public profiles. The scraper does accept an optional li_at as a fallback for profiles that are not publicly visible, but you never have to provide one, and the default path does not use it.

Is scraping public LinkedIn profiles legal?
In hiQ Labs v. LinkedIn, US courts held that scraping publicly available profile data does not violate the Computer Fraud and Abuse Act. That is not legal advice, and it is not the whole picture: LinkedIn's Terms of Service still prohibit scraping, and profile data is personal data under GDPR, which means you need a lawful basis to process it and must honour deletion requests. Consult a lawyer for your specific use case.

What happens with private profiles?
You get a record with a status code and no profile fields, rather than a silent failure. Filter on status before processing.

Can I search for profiles rather than supply URLs?
Yes — use the LinkedIn Search Scraper to find profiles by keyword, then feed those URLs into this one.

How fast is it?
Concurrency is configurable. Keep max_concurrency low (1–3) for large batches; LinkedIn tolerates a steady trickle far better than a burst.

Try it: LinkedIn Profile Scraper on Apify Store — no cookies, no account, pay per profile.

How to Scrape All Posts From a Facebook Page (No Cookies)

Mena489 — Mon, 24 Aug 2026 19:57:47 +0000

Pulling every post from a Facebook page — with reactions, comments, shares and video views on each — is the fastest way to see what a competitor is actually doing rather than what they say they are doing. This guide shows how to do it without a Facebook account.

Short answer: give the Facebook Profile/Page Scraper a page URL and a post limit, and it walks the page's timeline and returns each post as structured JSON.

What you can and cannot get

Be clear on the boundary before you start:

You can scrape any public page or public profile — business pages, creator pages, public figures, and personal profiles set to public. This is the same content any logged-out visitor sees.

You cannot scrape private profiles, friends-only posts, private groups, or anything behind a login. Nor should you try; that is unauthorised access, not scraping.

For competitor research, brand monitoring and creator analytics, the public surface is where the useful data lives anyway.

Step 1 — Point it at a page

{
    "url": "https://www.facebook.com/OrangeEgyptOfficial",
    "limit": 50,
    "startDate": "2026-01-01",
    "endDate": "2026-06-30"
}

limit caps how many posts come back — start small while you check the output shape, then raise it. startDate and endDate are optional; supply them to restrict the run to a window rather than paying to walk a page's entire history.

Step 2 — Read the results

Each post arrives as one record:

{
    "post_id": "122150535824513880",
    "type": "video",
    "permalink_url": "https://www.facebook.com/reel/625885050263599/",
    "owner_name": "El fútbol",
    "thumbnail": "https://scontent.fdxb3-4.fna.fbcdn.net/...",
    "statistics": {
        "reaction_count": { "count": 1204, "is_empty": false },
        "comment_count": 87,
        "shares_count": 33,
        "video_view_count": 41208
    }
}

The type field (video, photo, text, reel) is what makes this data useful. Grouping engagement by post type tells you which format a page's audience actually responds to — usually a very different answer from what the page publishes most of.

One detail that trips people up: reaction_count is an object, not a number, and is_empty distinguishes a genuine zero from a count Facebook declined to disclose. Averaging count without checking is_empty silently drags your engagement numbers down.

Turning posts into a content strategy

Three queries answer most competitive-research questions:

Which format wins?

from collections import defaultdict

by_type = defaultdict(list)
for post in posts:
    stats = post.get("statistics", {})
    reactions = stats.get("reaction_count", {})
    if not reactions.get("is_empty"):
        by_type[post["type"]].append(reactions.get("count", 0))

for post_type, counts in by_type.items():
    print(post_type, "avg reactions:", sum(counts) // max(len(counts), 1))

What are their top 10 posts? Sort by reactions and read the captions. That is your content brief.

How often do they post? Count posts per week from the date window. Posting cadence is the cheapest competitive metric there is, and most brands guess it wrong.

Calling it from your own code

Python

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("mina_safwat/facebook-profile-scraper").call(
    run_input={"url": "https://www.facebook.com/OrangeEgyptOfficial", "limit": 50}
)

posts = list(client.dataset(run["defaultDatasetId"]).iterate_items())
top = sorted(posts, key=lambda p: p["statistics"]["reaction_count"].get("count", 0), reverse=True)
for post in top[:10]:
    print(post["statistics"]["reaction_count"]["count"], post["permalink_url"])

cURL

curl -X POST "https://api.apify.com/v2/acts/mina_safwat~facebook-profile-scraper/run-sync-get-dataset-items?token=<YOUR_APIFY_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://www.facebook.com/OrangeEgyptOfficial","limit":25}'

What it costs

$0.0095 per post returned. Pulling a competitor's last 500 posts costs under $5 — roughly the price of a coffee for a complete picture of a year of their content strategy.

Common use cases

Competitive content audits. Pull three rivals' last 200 posts and compare format mix, cadence and engagement side by side.
Influencer vetting. Check whether a creator's engagement is consistent or spiked by a single viral post.
Campaign tracking. Monitor a brand's posts through a launch window and measure lift.
Historical archives. Preserve a page's public output before it changes or disappears.

FAQ

Can I scrape a personal profile?
Only if it is set to public. Private profiles return no post data.

How far back can it go?
As far as the page's public timeline extends. Use startDate to bound the run rather than paying to walk years of history you will not use.

Do I get comment text?
No — you get comment counts. For the comments themselves, run the Facebook Comments Scraper on the post URLs this returns.

What if I only need one specific post?
Use the Facebook Post Scraper instead — it takes post URLs directly and is cheaper per record.

Is this legal?
Scraping publicly accessible pages is generally permitted in many jurisdictions, but Facebook's Terms of Service prohibit automated collection, and post data tied to identifiable people is personal data under GDPR. This is not legal advice — confirm your basis before collecting at scale, and do not scrape private content.

Try it: Facebook Profile/Page Scraper on Apify Store — no cookies, $0.0095 per post.

How to Scrape Facebook Posts Without Login or Cookies (2026 Guide)

Mena489 — Mon, 24 Aug 2026 19:56:06 +0000

You can scrape any public Facebook post — reactions, comments, shares, view counts, media and author details — without a Facebook account, a session cookie, or a browser. This guide shows the manual approach, why it breaks, and the one-call method that takes about two seconds per post.

Short answer: send the post URL to the Facebook Post Scraper and read the structured JSON it returns. No login, no cookies, no browser automation.

Why scraping Facebook posts is harder than it looks

Facebook stopped being a scrapable HTML site years ago. Three things break naive scrapers:

1. The page you get is not the page you see. A logged-out request to a post URL usually returns a shell — a few kilobytes of markup with an empty <div id="mount"> and no post content. The real data arrives later over a separate GraphQL call the page makes to itself. If your scraper parses the first response, it parses nothing. Worse, it returns HTTP 200, so naive error handling reports success.

2. Every route is token-gated. Facebook's internal GraphQL endpoint requires an lsd token, a fb_dtsg token, and a doc ID that rotate constantly. You cannot hardcode them. They have to be minted from a live page load and replayed within their validity window.

3. Logged-out routes disappear without notice. Routes that worked last month return 404 today. Anything you build against a specific URL pattern has a short shelf life.

The usual answers — Puppeteer with a logged-in account, or buying session cookies — both fail commercially. A logged-in account scraping at volume gets checkpointed within hours, and cookie-based scraping means your data pipeline dies the moment that account is flagged.

The reliable approach: no account at all

The Facebook Post Scraper works entirely from public, logged-out data. It mints the tokens it needs per request, replays them against the right internal endpoint, and returns parsed JSON. Because it never authenticates, there is no account to ban and no cookie to expire.

Step 1 — Give it post URLs

{
    "urls": [
        { "url": "https://www.facebook.com/apifytech/posts/pfbid0ABC123" },
        { "url": "https://www.facebook.com/watch/?v=1234567890" },
        { "url": "https://www.facebook.com/reel/625885050263599/" }
    ],
    "maxConcurrency": 5
}

Posts, videos, reels and photo posts all work. Pass one URL or ten thousand.

Step 2 — Read the results

Each post comes back as one dataset record:

{
    "post_id": "122150535824513880",
    "type": "video",
    "permalink_url": "https://www.facebook.com/reel/625885050263599/",
    "owner_name": "El fútbol",
    "thumbnail": "https://scontent.fdxb3-4.fna.fbcdn.net/...",
    "statistics": {
        "reaction_count": { "count": 1204, "is_empty": false },
        "comment_count": 87,
        "shares_count": 33,
        "video_view_count": 41208
    }
}

Note reaction_count.is_empty. Facebook returns a zero-valued object both when a post genuinely has no reactions and when it withholds the count. The is_empty flag is the only way to tell "zero" from "not disclosed" — treat a missing flag as unknown rather than as zero, or your engagement averages will be quietly wrong.

Posts that were deleted or made private return an explicit error instead of silent nulls:

{
    "url": "https://www.facebook.com/apifytech/posts/deleted",
    "data": { "permalink_url": "https://www.facebook.com/apifytech/posts/deleted" },
    "error": "not found"
}

Calling it from your own code

The Apify API runs the scraper and returns the dataset in a single request.

cURL

curl -X POST "https://api.apify.com/v2/acts/mina_safwat~facebook-post-scraper-ppr/run-sync-get-dataset-items?token=<YOUR_APIFY_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"urls":[{"url":"https://www.facebook.com/apifytech/posts/pfbid0ABC123"}]}'

Python

from apify_client import ApifyClient

client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("mina_safwat/facebook-post-scraper-ppr").call(
    run_input={"urls": [{"url": "https://www.facebook.com/apifytech/posts/pfbid0ABC123"}]}
)

for post in client.dataset(run["defaultDatasetId"]).iterate_items():
    stats = post.get("statistics", {})
    print(post["post_id"], stats.get("comment_count"), stats.get("shares_count"))

JavaScript

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: '<YOUR_APIFY_TOKEN>' });
const run = await client.actor('mina_safwat/facebook-post-scraper-ppr').call({
    urls: [{ url: 'https://www.facebook.com/apifytech/posts/pfbid0ABC123' }],
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Results also export as JSON, CSV, Excel or XML from the Apify Console if you would rather not write code at all.

What it costs

Pricing is per post returned, and drops with your Apify plan: $0.009 per post on the free plan, $0.007 on Bronze, $0.005 on Silver, and $0.003 on Gold and above. Failed or deleted posts are not charged as results.

Scraping 10,000 posts costs roughly $30 on Gold — cheaper than a single hour of engineering time spent keeping a Puppeteer fleet alive.

Common use cases

Competitor benchmarking. Track reactions, comments and shares across a rival's posts to see which formats actually land.
Influencer verification. Check whether engagement on a creator's posts matches their claimed reach before you pay them.
Ad creative research. Pull the top-performing organic posts in your niche and reverse-engineer what works.
Brand monitoring. Watch engagement on posts that mention you and catch a problem while it is still small.

FAQ

Do I need a Facebook account to scrape posts?
No. This scraper reads only public, logged-out data. There is no account to create, no password to store, and no account that can be banned.

Is scraping public Facebook posts legal?
Scraping publicly accessible data is generally permitted in many jurisdictions, and courts in the US have repeatedly declined to treat public-web scraping as unauthorised access. That is not legal advice. You remain responsible for complying with Facebook's Terms of Service, with GDPR and similar laws where personal data is involved, and with the rules of your own jurisdiction. Do not scrape private posts or use the data to build profiles of private individuals.

Can it scrape comments too?
This scraper returns comment counts. For the comment text itself, use the Facebook Comments Scraper.

Can it scrape every post from a page rather than one URL at a time?
Yes — that is a different tool. Use the Facebook Profile/Page Scraper, which walks a page's whole timeline.

How fast is it?
Roughly two seconds per post, and it runs them concurrently. Ten thousand posts finish in well under an hour.

Why not just use the official Graph API?
The Graph API only returns posts from pages you administer, and requires app review for almost anything useful. For competitor and market research it is not an option.

Try it: Facebook Post Scraper on Apify Store — no login, no cookies, pay only for posts returned.