Brand Mentions Monitor — Web Mentions & Sentiment
Pricing
Pay per event + usage
Track brand mentions on news, blogs, forums & shops with sentiment, emotions, domain rank & monthly trend. Only-new mode for schedules.
Brand names, product names or phrases to monitor, one per line (up to 50). Phrases with spaces are matched exactly (turn off "Exact phrase" below to match the words anywhere). Matching is case-insensitive. A keyword whose search delivers no mention (nothing matches, or nothing new) is charged one "Search without delivered mentions" ($0.0303), because the search is paid for either way.
How many mentions to deliver per keyword (1–20,000). Each 1,000 mentions are one search.
Only mentions published on or after this date. Use 2025-01-31 or a relative period like "30 days", "6 months" or "1 year". Pages without a known publication date are left out when a date range is set (see "Date range applies to").
Only mentions published on or before this date (inclusive). Same formats as "From date".
Only mentions written in these languages, as ISO codes (en, de, es, fr, tr, …). Empty = all languages.
Only mentions from websites in these countries, as ISO codes (US, GB, DE, TR, …) or English names. Empty = all countries.
Remember what was already delivered and return only mentions that are new since the previous run with the same keyword and filters. The first run returns everything (up to the maximum). Repeats are never charged as mentions, but each search is paid for: a run where nothing is new costs $0.0303 per keyword ("Search without delivered mentions"), and every returned mention that is skipped as already delivered costs $0.0001 ("Fetched mention not delivered").
Match multi-word keywords as an exact phrase ("apple watch"). Turn off to find pages with the words in any order. You can also put your own quotes around a keyword.
Require the keyword in specific parts of the page. Empty = anywhere in the indexed text.
"All mentions" returns every matching text block. "One per domain" returns only the best mention from each website — useful to see how many different sites talk about you.
Order of the results. With "Only new mentions" the default is "Recently discovered".
"Publication date" uses the date the page was published (pages without a known date are excluded). "Discovery date" uses when the page was found by our crawler — every page has one.
Only mentions on these websites (e.g. reddit.com, nytimes.com). Subdomains are included.
Scale for domain rank and page rank, in the output and in the rank filters below.
Only websites with at least this authority (on the rank scale above).
Only websites with at most this authority.
Only pages with at least this authority (on the rank scale above).
Only pages with at most this authority.
Skip websites with a higher spam score (0–100). Try 30 to remove most low-quality sites.
Only text blocks with at least this content quality score (0–100).
Only mentions with this sentiment (probability at least "Minimum sentiment probability").
0–1, used with "Sentiment". A mention counts as positive/negative/neutral when that probability is at least this value. Default 0.4.
Only mentions expressing this emotion (probability at least "Minimum emotion probability").
0–1, used with "Emotion". A mention counts toward the emotion when that probability is at least this value. Default 0.4.
Optional name for the "Only new mentions" memory (e.g. "acme-weekly"). Runs with the same name share one memory per keyword, even if you change filters. Empty = one memory per keyword + filter combination.
Forget previously delivered mentions for this run's keywords and treat everything as new.
Adds to each keyword's summary: the total number of matching mentions on the whole web with sentiment and emotion counts, top domains, countries, languages, categories and a trend over time — not just for the mentions you downloaded. $0.08 once per keyword when data is returned; $0.0786 when the lookup was made but returned no data (skipped for free when the mentions search already found no matching page). The free summary of delivered mentions is always included.
Time buckets of the web-wide trend. The trend covers the date range above, or the last 12 months (max 400 buckets).
How many top domains, categories, countries etc. the web-wide statistics list (1–20). Also the length of the top-domain list in the free summary.
0–1. Minimum probability for a mention to count as positive/negative/neutral in the web-wide statistics. Default 0.4.
0–1. Minimum probability for a mention to count toward an emotion in the web-wide statistics. Default 0.4.
Expert filter expression added to the filters above with AND. Format: [field, operator, value] or [[…], "and"/"or", […]]. Fields: url, domain, main_domain, country, language, url_rank, spam_score, domain_rank, score, fetch_time, page_types, page_category, content_info.title, content_info.main_title, content_info.snippet, content_info.author, content_info.content_type, content_info.date_published, content_info.content_quality_score, content_info.level, content_info.text_category, content_info.connotation_types.positive|negative|neutral, content_info.sentiment_connotations.anger|happiness|love|sadness|share|fun. Operators: =, <>, <, <=, >, >=, in, not_in, like, not_like, regex, not_regex, match, not_match, has, has_not. Dates look like "2025-01-31 00:00:00 +00:00". Max 8 conditions in total.
