Facebook Group Post Search — Exact Mentions

Find exact words and phrases in publicly discoverable Facebook group posts. Collect matching post text, URLs, group and author details, publication dates and available engagement counts. Use exclusions to remove unwanted mentions and inspect matchedKeywords to see why each post was included. No Facebook login, cookies or AI key required.

Pricing: $0.029 per run + $0.00199 per matched post delivered ($1.99 per 1,000 posts, plus the run fee). Platform usage is included. The run fee applies even when no posts match. Results and complete coverage are not guaranteed.

When to use it

• Find literal mentions of a brand, product, place or phrase in discoverable public group posts.
• Collect research samples whose matching rule can be checked against the original text.
• Exclude unwanted phrases without sending post text to an AI classifier.

For example, pasta madre matches that phrase with normalized spacing and capitalization. It does not expand to lievito naturale. Add multiple phrases when you want alternatives. A negative sentence mentioning your phrase is still a literal mention; this Actor does not infer sentiment or purchase intent.

Quick start

1. Enter one or more words or phrases.
2. Choose a maximum number of posts per phrase, a publication period and search market.
3. Start the run and review Matching posts and Search completion and limits.

Start with a small sample:

{"keywords":["python programming"],"maxPosts":3,"afterDate":"anytime","countryCode":"us"}

The maximum is a target, not a promise. This phrase has produced useful posts in private tests, but yield and runtime vary. A narrow phrase can return fewer posts than requested, including zero.

Matching and filters

Input	Behavior
keywords	Match any listed word or phrase in the available original post text; up to 100 phrases. No synonyms or semantic expansion.
excludeKeywords	Reject a post containing any excluded word or phrase, even when an inclusion matches.
maxPosts	Per-phrase upper bound. Global deduplication, availability and search limits can reduce the count. Large targets do not guarantee large datasets.
afterDate	Use anytime, last_week (7 days), last_month (30 days), this_year, a start date such as 2026-01-01, or 14 days. All dates use UTC. Missing dates are excluded when filtering is active.
countryCode	Discovery market: US, UK, Italy, Germany, France or Spain. This does not verify where authors live.

Matching ignores letter case and normalizes Unicode and whitespace. Accents and punctuation remain meaningful. Word boundaries prevent apple from matching inside pineapple. Regular expressions are not supported. Missing post text cannot match; search snippets are never substituted for the post text.

Each post is saved once per run. When it matches several configured phrases, matchedKeywords lists all matches and searchKeyword records the first matching phrase with remaining capacity. Deduplication does not extend across separate runs.

Output

Fields	Meaning
post_id, post_url, text	Unique post identity, source link and available public text. Text may be a preview; completeness is not guaranteed.
matchedKeywords, matchReason	Literal phrases found and the rule explaining inclusion. No semantic probability is implied.
group_id, group_name, group_url	Available group context.
author_id, author_name, author_url	Available author details.
date, processed_at	Publication time when available and collection time.
reaction_count, comment_count, share_count	Public engagement counts at collection time.
image_url	An available post image; not a complete media export.
searchKeyword, searchCountry, textCompleteness	Attribution, discovery market and availability metadata.

Unavailable fields are null. The dataset contains posts, not comment text, member lists or private-group content. Export the dataset using Apify's JSON, CSV or Excel options. API/MCP consumers should inspect the linked summary as well as the dataset.

Example output

Illustrative record with fictional identifiers; unavailable source fields stay null:

{
  "post_id": "123456789",
  "post_url": "https://www.facebook.com/groups/example/posts/123456789/",
  "text": "Looking for Python programming study partners.",
  "matchedKeywords": ["python programming"],
  "matchReason": "literal-phrase",
  "searchKeyword": "python programming",
  "searchCountry": "us",
  "textCompleteness": "unknown",
  "date": null,
  "reaction_count": null
}

Understand incomplete and empty results

Discovery is bounded. It stops when targets are reached, the source ends, a spending limit is reached or the operating-cost/time/request limits prevent further searching. This is not a complete group archive or an exhaustive search of Facebook.

The SUMMARY record distinguishes:

• SUCCEEDED: targets reached or recognized source pagination ended. An empty dataset can be valid.
• PARTIAL: searching stopped at a limit; some phrases may be not-started. Saved posts remain useful, but more matching posts may exist.
• FAILED: a source, processing or storage error prevented completion. Already saved records remain available.

A platform run marked SUCCEEDED can have a PARTIAL search summary. economic-cost-limit is an operating-cost stop, customer-spending-limit is your budget, and source-exhausted means the recognized source pagination ended. Rejected, excluded, unavailable and out-of-range counts explain why candidates were not delivered.

Pricing and comparison

One run at the supported 256 MB memory costs $0.029 + $0.00199 × delivered posts, on every Apify tier. There is no result charge for rejected posts or duplicates. The start fee remains payable for empty or failed searches. Set a maximum run charge; the minimum supported limit is $0.03099, enough for the start and one result.

Delivered posts	This Actor, all tiers	Lofomachines Free	Lofomachines Gold/Platinum/Diamond
0	$0.02900	$0.03000	$0.03000
1	$0.03099	$0.03499	$0.03299
10	$0.04890	$0.07990	$0.05990
100	$0.22800	$0.52900	$0.32900

Tariff comparison checked 3 November 2026 against the Lofomachines pricing, assuming one run and one start event for each. These are calculations for the same delivered count, not promises that either tool will find that count or the same posts. The 100-post example is a price calculation, not a verified yield. Intermediate Lofomachines tiers fall between the shown prices.

Lofomachines advertises semantic topic relevance; this product verifies literal mentions and exclusions. Choose based on the matching behavior you need. A lower tariff is not evidence of equal semantic coverage or a universal lowest-price claim.

Which tool fits your search?

Need	This Actor	Lofomachines (documented behavior)
Literal phrase evidence	matchedKeywords and matchReason explain every saved match	Topic relevance with an estimated probability
Meaning and synonyms	Not supported; list each phrase explicitly	Semantic matching advertised
Exclude unwanted phrases	Global literal exclusions, applied before saving	An exclusion input is not documented in the reviewed listing
Access	Public, anonymously readable group posts	Public posts without login or cookies
Dates	Presets, UTC start dates and relative day counts	Presets plus custom API dates advertised
Coverage	Bounded discovery; incomplete searches are identified in SUMMARY	Limited public discovery; maximum counts are not guarantees
Price	$0.029/run + $1.99/1,000 delivered posts	$0.03/run + $2.99–$4.99/1,000 results, depending on tier

Comparison checked 3 November 2026 against the Lofomachines listing. Their column describes published claims, not independently verified performance. Our literal matching and exclusions have been tested; semantic coverage is deliberately different. Neither column implies complete Facebook coverage.

FAQ

Does it use AI? No. Inclusion and exclusion use deterministic matching against available public post text. No AI account or enrichment service is required.

Can it access private groups or login-only posts? No. It only processes anonymously readable posts found through public discovery.

Why did it return fewer posts than requested? Some candidates were excluded, lacked usable text, were outside the date range, were duplicates or were inaccessible. Search and cost limits can also stop the run. Check the summary before interpreting low counts as low real-world activity.

Are the posts ordered by date? No chronological ordering is promised. Sort the available date field after export.

Will it find every mention of my brand? No. It searches publicly discoverable group posts and cannot guarantee complete indexing, access or text. Exact matching also does not recognize synonyms, misspellings or unlisted variations.

Does the market prove author location? No. It tunes discovery, not residency or exact geolocation.

Can I resume a stopped run? Automatic restart/resurrection is not supported by this Actor. Existing results are preserved; it refuses to repeat work on a restarted run to avoid duplicate billing. A new run is separate and may find overlapping posts.

Ready-to-run examples

Use these inputs to try a small sample. Set a maximum run charge of $0.10; the start fee and delivered posts determine the actual charge.

Example	Input	Expected maximum
Sourdough mentions	keywords: ["sourdough"], afterDate: "anytime", countryCode: "us", maxPosts: 5	5 posts
Python this year	keywords: ["python programming"], afterDate: "this_year", countryCode: "us", maxPosts: 3	3 posts
Italian baking phrases	keywords: ["pasta madre", "lievito madre"], excludeKeywords: ["vendita"], afterDate: "anytime", countryCode: "it", maxPosts: 1	2 posts across both phrases

Actual results can be lower or empty. These examples passed private validation; availability changes over time.

More public-source research

For video research, Bilibili Scraper collects Bilibili video, creator and comment data. It is a separate Actor with its own pricing and source coverage.