TikTok Shop Market Intelligence Suite

Create one evidence-backed TikTok Shop market report covering products, sampled sellers, public creators, video creatives, and review sentiment.

The Actor uses its own bounded HTTP and browser pipeline. It does not call external data APIs or other paid Actors, so one Suite run does not create hidden double billing.

What the report answers

• Which captured products have the strongest public demand, rating, price, and evidence signals?
• Which sampled sellers lead within the captured product set?
• Which public creator and video leads are worth reviewing?
• Which hooks, calls to action, formats, and creative angles appear in public captions?
• What do captured public text reviews say, and how large is the sample?
• Where is coverage incomplete or evidence too weak for a confident conclusion?

Analysis modes

• FULL: products, sampled sellers, creators, creatives, and public review sentiment.
• MARKET: products, sampled sellers, creators, and creatives without review-page extraction.
• PRODUCTS: the lightest product and sampled-seller report.

Input fields

Provide known TikTok Shop product URLs, specific product niches, or both. The prefilled example analyzes one HPBS product in FULL mode, with no keyword search, up to three reviews and three creatives, a 20-page cap and a 240-second total budget. Larger studies can increase these limits and the Apify run timeout. Missing public reviews or videos remain explicitly reported.

Output contract

One dataset row represents one market intelligence report. It contains:

• top-level counts and top opportunities for fast table review;
• rankedProducts;
• sellers based only on the products captured in this report;
• creators and creatives with explicit association states;
• sentimentAnalyses with review sample sizes and topics;
• quality, coverage, and evidenceNotes;
• a pipeline declaration showing zero external data APIs and zero external Actors.

Missing public metrics remain null. Search-context videos are leads, not confirmed affiliate relationships. Public sold counts are cumulative signals unless TikTok explicitly exposes a fixed-period value.

MCP and agent usage

Use this Actor when a user requests a combined TikTok Shop market study rather than one narrow extraction. For a single outcome, prefer the specialized NexaScout Actor. This keeps agent selection predictable and avoids unnecessarily expensive FULL reports.

Local verification

Run npm ci, npm test, apify validate-schema, and apify push. Node.js 22 or newer is recommended and used by the Apify image.

Time limits and intermediate evidence

The run stops collection 25 seconds before the earlier of maxRunSeconds or the platform timeout, allowing active browser work to stop and a report to be saved. Startup time counts when supplied by Apify. Network requests share that deadline. This reserve is not a guarantee against platform outages or slow storage.

Products, videos and reviews are saved during extraction to the default key-value store under CHECKPOINT. This is intermediate evidence, not a completed or automatically resumed report. It stays available if the process is interrupted; CHECKPOINT_STATUS records normal completion. It does not add dataset rows or trigger report billing events.

The final dataset contains at most one unified report, also saved as REPORT. OUTPUT.coverageState = TIME_BUDGET_LIMITED and qualityState = PARTIAL_VERIFIED explicitly identify a time-limited report. A missing section must not be read as zero market activity. No report is fabricated when no usable products were captured.

Existing saved tasks retain their old inputs. To use the compact example, update the saved task or paste INPUT.example.json. Parsing, relevance filters, association evidence and scoring are unchanged in this release.