Facebook Page Post Tracker: AI Winning Posts & Offers

New winning Facebook posts and offers from pages you track, in your inbox weekly.

⚡ First 25 posts free · 💵 $0.0043 per qualified post · 🤖 AI ≈ $0.030 per post all in, on plans with model access or your key · ⏱️ demo in seconds

Try it in 30 seconds. Click Try it — the form is pre-filled with 8 sample posts from 3 pages; 4 qualify in seconds, nothing charged. Paste page links or pick a preset to search live; the first 25 posts are free. → What a run costs · What the AI adds · Run it weekly

⚡ At a glance

What you get	Only the posts worth your time — winners vs their page's own median, offers, launches, complaints — ranked, with the reason and where to reach the poster
You provide	3–10 public Facebook page links — or only a goal, and keywords are derived from it
Output	a ranked sheet, best first — Excel, CSV or Google Sheets — plus an e-mail digest of what is new each run
Typical run	4 pages × 20 posts = 80 scanned → ≈ 28 qualified (≈ 35 %, estimate until a live run is measured)
Cost of that run	$0.60 rules only (≈ $0.021 per post all in) · $0.84 with AI (≈ $0.030 per post all in) — the $0.48 source scan on your account included
Free tier	first 25 qualified posts per account
Keys / setup	AI on your plan: with Apify model access, every post, no key · any other plan: your first 5 real posts live on us, then a free Gemini / Groq key = 1–3 AI verdicts per run, a paid key every post
Languages	Free rules read English, Spanish, Portuguese, French and German; other languages: leave the AI on — it reads any language
Which plan	Free demo and your own list (list mode, template): any plan · live search: a plan that runs Store Actors
Works with	Schedules, webhooks, Sheets, Slack, Teams, Google Chat, Discord, Telegram, Make / n8n / Zapier, MCP & AI agents

🎯 What this Actor does

Only the posts that pass your filters are charged:

• A fair winner test. Every post against its own page's median engagement: "5.6× this page's median".
• What kind of post it is. Promotion, launch, hiring, event, complaint, question or buying intent — confirmed by quotes with AI.
• The offer, extracted. Discount, price, deadline and promo code, as written.
• A reply and a way to reach them. A drafted reply for buying-intent posts; every row links the poster's page.
• Which page is winning. A page leaderboard in every report; opt-in running ads and niche insights.
• Only new posts. Never delivered or charged twice.

🏆 Find top performing Facebook posts

Outlier multiple (default 2×), "Only outlier posts", minimum reactions and a post window — free filters, applied before any charge.

🔭 Monitor competitor Facebook pages

Paste page links (20 posts per page, ≈ $0.006 each on your account) and schedule it weekly: from week 2 only new posts are read.

📊 Facebook competitor analysis: leaderboard, ads, what wins

Every REPORT ranks the tracked pages: posts per week, median engagement, winners, offers, change since last run. Opt-in: each page's top 3 running ads ($0.10 per page with ads) and what wins in your niche — media, weekday, hour, length, type ($0.05 a run).

🛒 Spot people asking where to buy (test)

Type keywords (e.g. "need a wedding photographer"): posts asking where to buy get buying_intent and a reply draft. The search returns mostly page posts (no groups or comments); the Buying intent preset caps its first scan at $0.15.

🚀 How to use it

1. Click Try it — the form is pre-filled with the free demo on sample posts; it works as is.
2. Type your page links and what you are looking for — they replace the sample and start the live search, capped by "Max source spend". Leave "AI assessment" on: your first 5 posts get live AI verdicts on any plan; all 25 free ones get them with Apify model access or a paid own key (Keys / setup).
3. Press Start and open the dataset — qualified posts land best-first; export them or wire a webhook.
4. Schedule it weekly (Actions → Schedule) with your address in "E-mail me the new posts" — only new posts are delivered and charged.

📦 Already using a Facebook posts scraper? Rank your Facebook scraper data

Paste the dataset ID of your last Facebook posts scraper run into Dataset ID — no second scrape, rules only by default, you pay only for the qualified posts (after every scrape).

📚 Learn more: Tutorial — track competitor Facebook pages step by step

🆚 Instead of a raw Facebook profile posts scraper?

That scraper returns raw rows; the median, sorting and offer are yours to do. Of the demo's 8 posts, 4 were dropped before any charge here (a live funnel follows 2026-11-01).

	Facebook Profile & Posts Scraper (cleansyntax/facebook-profile-posts-scraper)	This Actor
Your time per 100 posts	≈ 10 h sorting (our estimate, 6 min each, not timed)	0 h — ranked, typed, explained
What you pay for	every row scraped, empty and duplicate ones too	only qualified posts — the rest costs $0 here
Offers and intent	"common fields (may vary)"	post type, offer, reply draft
Repeat runs	the same posts again, billed again	only new posts read and charged
Same 100 qualified posts, all in	≈ $1.72 for ≈ 286 unsorted raw rows	$2.15 rules only ($0.43 here + the same $1.72 scan) · $3.02 with AI
Price	$0.006 per row · 435 users · 3 % of runs failed (30 days)	from $0.0043 per qualified post, first 25 free

vs other Facebook post scrapers and free tools

	Price	What it gives you	What is missing
This Actor	$0.0043 per qualified post · 100 posts ≈ $2.15 all in (AI ≈ $3.02)	page median, type, offer, reply, memory	comments (2026-11-05)
thedoor/facebook-page-scraper	$0.0025 per result · 279 users · ≈ $0.72 for 286 rows	page posts and comments, raw	page median, offers, memory
scraper_one/facebook-posts-scraper	$0.0025 per result + $0.0025 per run · 209 users · ≈ $0.72	posts, raw	page median, offers, memory
pratikdani/facebook-post-scraper	$0.025 per post + $0.002 per run · 127 users · 17 % of runs failed	one post per URL, raw	pages, ranking, offers
Meta Business Suite — Pages to Watch (free)	$0	official numbers for a few watched pages	no API, no scheduled export, no AI tagging

Prices, users / 30 days, failed runs: Store listings, 2026-10-05.

💵 Pricing — what a run really costs

Pay per result: one event per qualified post, plus one per tracked page on scheduled runs.

Event	When it is charged	Price
free-tier	your first 25 qualified posts on this Actor, in any mode	$0.00
qualified-post-basic	AI off — page baseline, post type, offer, rule score	$0.0043
qualified-post-ai	AI on — plus quotes, fit score and reason, reply draft, next step	$0.01
page-monitored	scheduled runs: per tracked page with a new post; first 8 free	$0.02
competitor-brief	opt-in includeAds: per page whose running ads joined the report (ad search capped at $0.30)	$0.10
niche-insights	opt-in insights: once per run with 10+ posts read and 3+ winners	$0.05

Never charged by us: posts below minScore, empty posts, reshares without text, posts outside your window, your exclude list, posts you already received, a page with nothing new. The source's scan bills your Apify account per post read, capped by maxDiscoveryChargeUsd.

How that compares — $0.0043 is 0.72× one raw row of the source scraper ($0.006) and level with the $0.00425 median row price of Facebook post scrapers; one qualified post replaces ≈ 2.9 raw rows ($0.0015 per raw-row equivalent), typed and ranked.

Worked example — 80 posts scanned → ≈ 28 qualified (≈ 35 %, estimate): source $0.48 on your account + 28 × $0.0043 = $0.60, about $0.021 per qualified post all in with rules only. With AI: $0.48 + 28 × $0.01 + ≈ $0.08 tokens = $0.84, about $0.030 per post all in. Your first 25 posts are free, so run 1 costs the $0.48 scan plus 3 × $0.0043.

At volume — a marketing team on 20 pages weekly (spend cap $3): week 1 ≈ $2.40 source; then ≈ $0.96 source + ≈ 50 posts × $0.0043 + 20 × $0.02 monitored pages ≈ $1.58 a week, ≈ $7 a month (estimate).

AI tier all in, per post: ≈ $0.013 ($0.01 + ≈ $0.003 tokens, estimate); ≈ $0.030 with the source scan.

🤖 What the AI tier adds

The same buying-intent post (“Can anyone recommend a good bike shop in Austin for a kids balance bike? Budget is around $120.”), AI off:

{"postType": "buying_intent", "matchedSignals": ["Can anyone recommend"], "ruleScore": 50, "label": "warm",
 "reason": "buying-intent post: “Can anyone recommend”", "replyDraft": null, "assessmentSource": "rules"}

AI on (qualified-post-ai, $0.01) — judged against your goal, with a reply; the real row of our own-key AI run of the example input, 2026-10-16, aiModel verbatim:

{"fitScore": 85, "score": 85, "label": "hot", "postType": "buying_intent",
 "matchedSignals": ["Can anyone recommend a good bike shop in Austin for a kids balance bike?", "Budget is around $120."],
 "reason": "Matches buying intent by asking for a good bike shop for a kids balance bike with a $120 budget.",
 "replyDraft": "Check out Trek Bicycle Austin or Bicycle Sport Shop—they both usually have a great selection of kids' balance bikes around that price range!",
 "assessmentSource": "ai", "aiModel": "google:gemini-3.5-flash-lite"}

Only the AI scores it against your goal, cites the post's words and writes the reply — about 6 minutes each by hand on 3 sample posts (our estimate, 2026-10-05; not yet timed).

On your plan (the same answer as Keys / setup): Apify model access → every post, no key; any other plan → your first 5 real posts get live AI verdicts on us, then a free Gemini key (aistudio.google.com/apikey) gives 1–3 per run, a paid key every post.

⚙️ Input

📤 Output

One item per qualified post. Views: Shortlist (best first), Offers & launches, Buying intent & replies (leads with "Reach them"), Engagement by page vs page median, Score breakdown. The Shortlist ranks hot → warm → cold, AI-assessed rows first, one score with Scored by (assessmentSource). Sort by score; ruleScore is the rule part and fitScore the AI's — threshold and trend across runs on ruleScore (the Score breakdown view has the parts).

Output fields

Upsert key for a CRM or Sheet: dedupeKey — id: plus the Facebook post id, one row per post across runs; itemId hashes the same key.

A DIGEST (new and hot counts, the change since the last run — "vs last run: qualified 41 → 55" — the top 10) and a client-ready REPORT page (OUTPUT.reportUrl: posts, page leaderboard, ads, insights) sit next to OUTPUT.

⭐ Found it useful? A review on the Store helps others find it.

🔁 Run it weekly

1. Open the saved task and click Schedule after your first real run (its status links it, OUTPUT.scheduleTaskUrl), with your address in notifyEmail: each run with new posts mails the digest and the trend vs the last run.
2. Leave dedupeAcrossRuns: true — the next run charges only what is new; give each client its own dedupeStoreName.
3. Paste a Slack, Teams, Google Chat or Discord webhook — or your Telegram bot's send link — into Send alerts to (one per line, detected by address) for the digest; a Make / n8n / Zapier hook there gets every new post as JSON and updates your tracker sheet on dedupeKey — no manual step.
4. From run 2, OUTPUT.history and the REPORT show every run; historyDatasetName keeps all posts in one dataset.

Weekly cost: from week 2 only new posts are read, ≈ 8 per page instead of 20. 4 pages → ≈ $0.19 source + ≈ 10 posts × $0.0043 + 4 × $0.02 monitored pages ≈ $0.31 a week (estimate), the first 8 monitored pages free.

🎯 Try it for your niche

Pick one under Start from a ready niche search and press Start; each first scan is capped at $0.15 of source spend.

Niche	What it looks for	Start
Competitor offers & launches	3 café / fast-food pages, promo and launch posts always kept	preset
Buying intent (test)	public posts asking for a wedding photographer (groups and comments not searched)	preset
Launch watch	3 sports-brand pages, outlier posts only	preset

🔌 Integrations, automation and API

After every cleansyntax/facebook-profile-posts-scraper run: in that Actor's Integrations tab add "Run succeeded → start this Actor" with sourceMode: dataset, datasetId = the finished run's resource.defaultDatasetId and enableAi: false — every scrape you already pay for is ranked, rules only.

Run outcomes — what your integration sees

• Webhook — {"event": "post.qualified", "post": {…}, "runId"} per post (posts.qualified per batch), a 429 / 5xx retried once; a POST that still fails is in OUTPUT.webhook.failedKeys, not re-sent. webhookSecret signs each body (X-Signature: sha256=…). HubSpot / Pipedrive: a Make / Zapier step, upsert on dedupeKey.
• Which call runs what — OUTPUT.mode: demo (never charged), live, list or dataset.
• AI agents / MCP — plain typed inputs; MCP server https://mcp.apify.com/?actors=rich_minds/fb-page-post-tracker. The snippets run the real search, capped; {} runs the free demo.

🧾 List-mode rows — what a pasted record needs

👥 Who is it for?

You are…	You run it to…	Start with
Social media manager tracking 5–30 competitor pages	see which rival posts won this week and their offers	preset Competitor offers & launches, weekly
Local business owner watching rivals	catch every rival discount, deadline and promo code	3–5 rival page links, post type Promotion
Growth marketer or founder testing demand	find public posts asking where to buy	preset Buying intent (test)
AI agent or n8n workflow	typed posts with scores, offers and reasons	the minimal agent input above

🧠 How the AI works

• Grounded. One post at a time: a type from a fixed list, quotes from the post, a fit score with a reason, the offer as written, a reply only for buying-intent and question posts.
• Your goal. The fit score measures the post against your goalDescription; only the best rule-scored candidates get the AI.
• No key where your Apify plan includes model access — otherwise your own key (llmProvider: "byok"). Default: anthropic/claude-haiku-4.5; an own Gemini key: google:gemini-3.5-flash-lite.
• Fallback. If the model fails, the post keeps its rule score, type and offer at the basic price.

🔒 Data, compliance and limits

• Public Facebook pages and posts only, read without a login. Follow Facebook's terms and applicable law (GDPR, CCPA); no unsolicited mass messaging.
• A count the source does not return stays null. A page needs 5 posts for its own baseline (fewer is weak_baseline; a scheduled run reuses the last median).
• Rule types are keyword-based in English, Spanish, Portuguese, French and German; the AI reads any language.

❓ FAQ

What will my first real search cost? 4 pages × 20 posts ≈ $0.48 on your account (capped at $0.50) and ≈ 28 qualified posts: the first 25 are free, the other 3 cost $0.0043 each.

How do I track competitor promotions? Paste the rivals' page links, pick Promotion under "Always qualify these post types", schedule it weekly: every discount, price, deadline and code lands in the Offers view.

Can it be my Facebook post engagement tracker? Yes — every post's engagement (reactions + 2 × comments + 3 × shares) is compared with its page's median, and the REPORT tracks each page from run to run.

Which Facebook competitor posts are worth copying? The ones far above their page's median and the ones naming an offer or a launch (Shortlist and Offers views); What wins in your niche names the pattern.

Do I still need the scraper it runs? No — it runs the source scraper on your account and adds the baseline, type, offer, AI and memory; list mode never runs it.

Why pay when Meta Business Suite is free? Its Pages to Watch is free, official and enough to eyeball a few pages — use it for that. It has no API, no scheduled export, no AI tagging, offers or reply drafts.

Will I be charged for the same post twice? No, with dedupeAcrossRuns on (default): every charged post is remembered by its post id. Adding a page or keyword keeps that memory; each dedupeStoreName and each postTypes / targetFlags filter set keeps its own — but a post already charged under one of them is delivered to the other as already-charged, never charged again.

🧩 More Actors from the same developer

Same buyers, same pay-per-result model:

• Facebook Ad Library Winning Ads Finder — Your competitors' longest-running Facebook ads — hook, offer and angle broken down, one row each.
• Instagram Competitor Content — Instagram posts that beat their account's median, in your inbox weekly.
• Twitter Lead Generation — People on X asking for a tool like yours — with a reply draft, every week.
• Website Change Monitor — Every real change on your competitors' pages, explained, in your inbox every week.
• YouTube Outlier Videos Finder — YouTube videos that beat their channel's usual views, scored against your brief with AI takeaways.

🆘 Support

Something missing or wrong? Open an issue on the Actor's page.

📝 Changelog

• 0.1.4 (2026-10-16) — One "Send alerts to" field takes your Slack, Discord, Teams, Google Chat, Telegram or Zapier / Make link, and the form is shorter; Your own Gemini key now gets AI verdicts without picking a model, and a post you already paid for under another watch list is delivered again but never charged twice; If your Apify plan can't run the source scraper, the run now ends successfully with a clear "use list mode" next step instead of failing; the page also shows how to rank a Facebook scraper dataset you already have, after every scrape.
• 0.1.3 (2026-10-09) — page leaderboard in every report; opt-in running ads and niche insights; pick lists for post types and flags; signed webhooks; es / pt / fr / de rules.
• 0.1.1–0.1.2 (2026-10-09) — initial release (winners vs each page's median, types, offers, replies, AI, memory, digest); weekly runs read only new posts; page-monitored.