Bid Industry Intelligence Analysis

Problem to Solve

Bidding data is scattered across platforms with inconsistent industry codes, making manual collection and analysis time-consuming and error-prone. Analysts need reliable market insights but face these pain points:
- Inconsistent data standards, leading to non-comparable statistics.
- Difficulty in quickly obtaining full-scale bidding data for specific industries, including quantity, amount, region, and participating enterprises.
- Inability to automatically distinguish between full statistics and sample analysis, risking overextrapolation of market share.
- Challenges in integrating bid and contract data to assess project lifecycles and enterprise relationships.

Core Workflow

This skill automates bidding industry intelligence analysis via API interfaces. Key steps include:

1. Industry Code Mapping: Use the AI industry search interface to map user-provided industry keywords (e.g., “medical devices”) to standardized first-, second-, and third-level industry codes (industryCode), ensuring consistent analysis standards.
2. Data Query and Statistics: Employ searchProjectApi to query quantities by overall or grouped conditions (e.g., time, region, purchase type) with pageNumber=0 to get total, then paginate for details (up to 50 per page). Data is deduplicated by id + publishTime.
3. Aggregation Analysis: Aggregate detailed data across dimensions like time, region, information category, and amount range. Calculate metrics such as project count, total amount (only for parseable amounts, reporting missing rate), and enterprise rankings (sorted by project count or amount).
4. Contract Analysis: Use searchProjectContactApi to fetch contract data, analyzing contract cycles, amounts, and party relationships for a full project lifecycle perspective.
5. Report Generation: Output a structured report including analysis scope, data coverage, core conclusions, trend tables, etc., with data evidence for each conclusion. Sample results must note sample size, pages, and cutoff time.

Applicability and Caveats

This skill is designed for bidding industry data analysis, but requires awareness of:
- Interface Call Limits: For bid searches, pageNumber=0 only returns total count, with a maximum of 50 per page; contract searches allow up to 100 per page. Full details require pagination by hasNext, recording actual retrieved counts.
- Data Completeness: When full data is not retrieved, the report must be labeled as “sample analysis,” not presented as complete market share. For missing amounts, report both the count of projects with amounts and the missing rate.
- Authentication and Quotas: Requires configuration of BBIAO_API_KEY and BBIAO_SERVER_URL; calls stop if unavailable. On insufficient balance, prompt user to recharge and retry only once.
- Analysis Standards: Different scopes (e.g., full statistics vs. paginated summaries) must not be directly compared; growth rate calculations require consistent comparison intervals, with absolute changes reported when the prior period is zero.
- Report Output: Avoid overextrapolation; clearly state data coverage and interface coverage gaps.

By following these steps and rules, this skill generates auditable market analysis reports, but users should understand its reliance on API data and standardized processing.