Skip to main content

Analytics

Every project has a Dashboard tab built from its mentions. All charts follow the filter bar (period, platform, sentiment, intent, type, media source, attribute, format, text; see Mentions), so a filter means the same thing on the dashboard, in the feed and in reports.

Hidden mentions and mentions marked irrelevant never count. Days are calendar days in the project's timezone.

Definitions​

MetricDefinition
MentionsPosts + comments + reviews that matched
InteractionsReactions + comments + shares
Social BuzzMentions + interactions
AuthorsDistinct authors (platform + handle)
Engagement rateEngagement divided by views; without views, engagement per follower per post
Net sentiment(positive − negative) ÷ labeled mentions

The dashboard​

KPI cards​

Total mentions, posts, comment mentions, review mentions, authors, sources, reactions, comments, shares, views, interactions and Social Buzz. Each card shows the change against the previous period (the same number of days just before). New means the previous period was zero.

Charts​

CardWhat it shows
Mentions & InteractionsPer day: mentions on the left axis, interactions on the right
Social Buzz & InteractionsBuzz per day
By platformMentions per platform with its change
Sentiment and Sentiment over timePositive, neutral, negative, by count and by buzz
Intent and Intent over timeProvider, seeker, general. Click one to filter
Own / Paid / EarnedSee Owned accounts
AttributesMentions per custom label. View attribute switches between attribute types
Interactions breakdownShare of reactions, comments and shares
Video lengthVideos by length, and which length gets the most interactions
Post formatsShorts, reels, lives, carousels, videos, images, text
Posts vs commentsThe split between posts and comments
Rating cardsAverage rating and its trend, when the project has reviews

Top mentions and Top influencers​

Top mentions lists the 20 mentions with the most interactions; Top influencers the 20 authors with the most mentions. Both have tabs for all, positive, negative and neutral. View all opens the feed or the influencer ranking with the same filters.

Topic cloud​

The phrases people use most (one to three words), colored by sentiment: green leans positive, red leans negative. It is built for Vietnamese: accents are kept, Vietnamese and English stopwords are dropped, and a hashtag is merged with the phrase it spells (#caphesuada counts as "cà phê sữa đá"). The project's own keywords and brand names are left out, because they appear in almost every post. Switch to Weighted by interactions (Social Buzz) to rank by engagement.

The top hashtags with their count in the previous period. A hashtag is rising when it is new (at least 2 mentions) or grew by 50% or more with at least 3 mentions.

Peaks​

Days with unusually many mentions (at least 3, and more than twice the median of the 14 days before, or a z-score of 2 or more), with the posts, authors, hashtags, sentiment and platforms that drove them. The 5 biggest are shown.

Growth​

Open Show growth at the bottom of the dashboard for views and engagement of every post at the end of each day, from collected snapshots, including older posts that are still growing. Snapshots are kept for 180 days.

Audience​

The Audience card estimates the audience in aggregate:

  • Gender, read from public display names as Vietnamese names (for example the middle names "Thị" and "Văn"). Ambiguous names, nicknames and brand or page names count as unknown.
  • Top provinces, from the profile location the platform returns (TikTok region, X location, YouTube channel country), mapped to Vietnam's 34 provinces, or "Abroad".

Every number is over distinct authors and comes with its sample size and coverage. Groups of fewer than 10 authors are hidden. Lantfly never stores or shows a label for an individual person, and does not estimate age by itself.

AI agent estimate. An AI agent connected through MCP can read the project, estimate gender and age bands in its own memory, and send back only the counts for a period (save-audience-estimate). The latest estimate shows in the Audience card with its method. The same rules apply: counts only, at least 10 authors, small groups hidden.

View settings​

View settings opens a read-only summary of what the dashboard is built from: platforms, tracking window, languages, industry, keywords, owned accounts, campaign creators, other sources and attributes with their rules.

Export​

Export PDF opens a print view of the dashboard with the current filters; print it or save it as PDF. For a frozen snapshot with a written summary, create a report.

Influencers​

On the Creators tab, the Influencers in the period view ranks the authors of the filtered mentions: mentions, posts, comments, interactions, shares, views, followers and sentiment, sortable by any column.

Each author has a sentiment score: (positive − negative) ÷ (positive + negative + neutral), from −1 to 1. Use Filters to pick platforms and a score range. Export downloads the ranking as CSV or XLSX (up to 10,000 rows).

Share of voice between brands​

Declare brands and their keywords on a project (brands in the API, up to 10). Project analytics then return the share of mentions and engagement of each brand, overall and per day.

Compare projects​

Listening → Compare puts 2 to 8 projects you can view side by side:

  • Side by side table: mentions, posts, comments, reactions, shares, interactions, views and sentiment, with each project's Share of mentions (share of voice);
  • share of voice by mentions and by interactions, per day and per platform;
  • Sources by platform;
  • topic clouds: the overlap between projects, shared topics only, or per project;
  • top mentions and top influencers across the projects;
  • audience estimates per project.

To compare one project over two periods, use Period A vs period B (for example against the Previous period or the Same period last year). Saved comparisons keep a set of projects for later.

Industry overview​

Social listening → Industry overview benchmarks every project of one industry you can open (your own, those assigned to you, and those shared with you):

  1. Pick an industry (the one set on each project).
  2. Choose the period, platforms and statuses.
  3. Read the overview: Social Buzz, interactions, mentions and comments against the previous period, the trend of the Top 10 projects, and a ranking table sortable by any column.

The sentiment score here is positive ÷ (positive + negative), from 0 to 1. At most the 200 newest projects are counted.

With the API​

EndpointReturns
GET /listening/projects/{project}/dashboardKPI cards with the previous period, series, sentiment, intent, media, attributes, durations, formats, reviews
GET /listening/projects/{project}/topicsTopic cloud (buzz=1, include_brand=1)
GET /listening/projects/{project}/hashtag-trendsTop hashtags with change and rising
GET /listening/projects/{project}/peaksSpike days
GET /listening/projects/{project}/influencersAuthor ranking (/export for CSV or XLSX)
GET /listening/projects/{project}/demographicsAggregate audience estimates
GET /listening/projects/{project}/audience-estimatesEstimates sent by AI agents (POST to send one)
GET /listening/compare?projects=a,bCompare 2–8 projects
GET /listening/overview?category=BeautyIndustry overview

All of them take the shared filters (from, to, platform, sentiment, intent, kind, media, attribute, q, author, ad, and more). Use compare_from and compare_to on the dashboard to pick another comparison period. See the Listening API.