2026 campaign-assessment score
How Pulsar scored
73 / 100
Pulsar pairs documented Chinese and international sources with speech and image analysis and a report-writing agent. Its creator-level baseline and TikTok collection in TRAC are undocumented.
| Criterion and evidence | Points |
|---|---|
Campaign explanationTRAC transcribes video from X, Instagram, YouTube and Facebook and adds image tags, logos and captions, and STORY's campaign debrief explains peaks and their causes. The listed transcription sources exclude Douyin and TikTok, and on-screen text in video is not described. | 21 / 30 |
Reaction analysisTRAC shows which topics, keywords and entities drive sentiment and emotion, and Themes clusters conversation with AI summaries. Its documentation does not separate reaction to a creator from reaction to the product or track buying intent, and Instagram comment text is limited to tagged mentions. | 14 / 20 |
Finished assessmentSTORY writes reports with a verdict, evidence, tested hypotheses, implications and caveats, with a campaign-debrief brief and scheduling. It has been in public beta since October 2026, and no complete sample report is public. | 16 / 20 |
Creator baselineCORE compares owned media with competitors across custom periods, and STORY's debrief compares a campaign with the seven days before launch. Its documentation does not compare a post with that creator's usual results. | 6 / 10 |
China and international coverageTRAC lists Douyin, Weibo, WeChat, Xiaohongshu, Bilibili, Kuaishou and Zhihu with collected fields and real-time-only limits, beside Instagram, YouTube, X and Facebook. Chinese sources are a premium bundle with no comment text in their field lists, and the TRAC source article omits TikTok. | 8 / 10 |
Source trailEvery STORY quote is a post's exact text linked to its source, and TRAC exports include a direct post URL column. The documented link covers quotes, and scores and charts have no equivalent description. | 8 / 10 |
Evidence differs by product. No product was tested head to head. Scoring methodology · Publisher disclaimer
Pulsar’s TRAC module does the research work. An analyst builds a search, and TRAC adds sentiment, emotion, topics and entities to what it collects. TRAC 101
TRAC transcribes video from X (original posts only), Instagram, YouTube and Facebook in 16 languages, Chinese among them, and indexes the transcripts like other text. Video Transcripts Image models add tags, logo detection and captions. Image Analysis
A February 2025 release added Weibo, Little Red Book, WeChat, Baidu, Douyin, Bilibili, Zhihu, Kuaishou and more than 680 Chinese news outlets and blogs, as a premium bundle. Chinese data sources Douyin returns posts, views and likes, in real time, with no history. TRAC Data Sources Data starts flowing within 48 hours of setup, so a team can follow a Douyin launch only from then on. It cannot go back, and comment text is absent from the listed fields. That list also omits TikTok, which Pulsar documents in CORE for a brand’s own footprint. TikTok in CORE
The newest piece is STORY, an agent that entered public beta in October 2026. Ask it why something happened and it names three to five candidate causes, tests them, and writes a verdict, evidence, implications and caveats. Reading a STORY report Every quote is a post’s exact text, linked to its source. STORY public beta Its campaign-debrief template compares the live period with the seven days before launch. Writing briefings Each creator’s usual performance is a separate comparison, and the help articles leave it out.
Pulsar fits teams running continuous social and media intelligence who want written reports built on those searches. unrandom covers a narrower job: reading a defined list of creator posts and their comments, and writing the assessment of that campaign. unrandom product