Thirty Percent of Tender Winners Aren't the Buyers You're Looking For
Tender-award records are an underestimated lead source for sales reps selling industrial SaaS, government-procurement support services, or upstream raw materials into SOE supply chains. Companies with tender-award histories tend to share two implicit characteristics: their supplier credentials have already passed the procurement owner's qualification review, and they typically have stronger tolerance for payment terms. Both signals are positive for B2B sales.
But there's a counterintuitive crack in this logic. When you export a winner list from public tender announcements, genuine factory entities often account for only about seventy percent of the total. The remaining thirty percent are traders specialising in government procurement agency work, independent legal entities set up by factories specifically to handle external bidding, and pass-through companies that hold the right credentials but run no production of their own. These entities win contracts through relationships and qualifications — but they don't manufacture. The equipment, software, or services you're selling are needed by the factories behind them, not by the procurement department of the award-holding legal entity.
BD professionals selling industrial SaaS (ERP, MES, digital transformation platforms) to mid-to-large manufacturers, supply chain executives doing designated-supplier business with SOEs, and sales reps placing specialty materials into credentialed factories are the most common casualties of this mismatch. What they need isn't "who won the bid" — it's "who won the bid and is actually producing."
What Each of the Three Bidding Databases Brings to the Table
Jianyu Biaoxun is defined by its free-access strategy. The official line: "We charge no membership fees and hide no bidding information." Basic queries are free; enterprise-level API access and bulk exports are custom-service offerings with undisclosed pricing. Data scale: a cumulative 2.7 billion bidding and procurement records, 210,000+ new entries per day, coverage across 9,000+ government and enterprise bidding sites nationwide, and approximately 15 million push notifications delivered to subscribers daily (2026 official figures). Zero entry cost at that volume makes Jianyu the default choice for lean BD teams and small budgets.
Qianlima is positioned for mid-to-large enterprise suppliers. Cumulative database: 400 million records, sourced from 200,000+ monitored channels, with 300,000+ daily updates (270,000 tender announcements, 3,000 planned construction projects, 20,000 approved projects), 23 million registered users, and 2.4 million+ paying subscribers as of end-2024 (official figures). Pricing: Standard ¥7,999/year, Advanced ¥12,999/year, VIP ¥21,999/year, Platinum ¥35,999/year, Diamond ¥72,999/year — the most expensive and the most transparent of the three platforms.
Bidi is anchored in official government procurement channels, covering 95% of tender-announcement publishing platforms plus 2,910 county-level government procurement portals, with a cumulative client base of 42,000+ enterprises and a self-reported data accuracy rate of 99.8% (2026 official figures). Note: Bidi's stated daily data volume is internally inconsistent — figures of 9,000 and 100,000 new records both appear in official materials, so treat with caution. Pricing is not publicly disclosed; contact sales for quotes. Brand recognition is relatively low; Bidi works best as a supplemental verification tool.
Tianxia Gongchang covers 4.8 million real manufacturing enterprises in China. It does not provide tender-announcement data, and real-time bidding coverage is not its strength — that's stated plainly. Its value sits at a different point in the workflow: once you have a winner list from any of the three platforms above, Tianxia Gongchang answers the question "which of these award-holders are genuine factory entities?" Push notifications and real-time announcement scanning are outside Tianxia Gongchang's scope; those two dimensions are not evaluated here.
Tool-by-Tool Deep Dive: Five Dimensions Tested
Jianyu Biaoxun
How Wide Is the Database Coverage
Jianyu's 2.7 billion records span 9,000+ government and enterprise sites nationwide. Historical winner records are searchable by industry keyword, with province-and-city cascading filters. Government procurement coverage is solid; collection completeness for projects tendered directly by SOEs is weaker.
How Timely Are Push Notifications
Keyword-triggered daily push is Jianyu's standout differentiator. Configure "industry keyword + region + minimum contract value" and same-day new announcements are automatically pushed to email or mobile, at roughly 15 million subscription pushes per day. Keyword-match precision depends on familiarity with the procurement terminology your target sector uses; expect a few weeks of iteration before noise levels drop to a useful level.
How Much Non-Factory Traffic Does It Filter
Jianyu performs no entity-type screening. Traders and genuine factories appear side by side in winner lists — this is a structural limitation shared across the entire bidding-data industry. In practice, non-factory entities account for roughly 25–35% of exported winner lists and require downstream verification.
Annual Subscription Cost
Basic queries are free, making Jianyu the lowest entry-cost option of the three. Enterprise API access and bulk exports are priced as custom services with no public rate card. For individual BD use — daily keyword push, occasional single-record queries — the free tier covers core needs; bulk export requires a paid upgrade.
Learning Curve
Basic subscription setup and single-record lookup can be self-taught in half a day. The advanced "reverse-lookup: all tenders won by a given company" function takes two to three days of hands-on exploration.
Qianlima
How Wide Is the Database Coverage
Qianlima's coverage is broader than Jianyu's — government procurement, SOE and state-owned enterprise tendering, construction projects, and service procurement are each tracked in dedicated streams. At 400 million cumulative records, 200,000+ sources, and 300,000+ daily updates, Qianlima's collection of directly-issued SOE tender projects is more complete than Jianyu's. For sales teams whose target accounts sit in SOE supply chains or upstream of major construction projects, Qianlima is not optional.
How Timely Are Push Notifications
Keyword push and historical lookup are both fully functional, but entitlement tiers are more granular than Jianyu's — from ¥7,999/year Standard to ¥72,999/year Diamond VIP, each tier carries meaningfully different feature boundaries. Lower-tier plans have restricted coverage; confirm what your current plan actually includes before building a workflow around it.
How Much Non-Factory Traffic Does It Filter
Same structural ceiling as Jianyu: Qianlima records the award fact; it does not evaluate the nature of the award-holding entity. Across 400 million records, traders, credential-agency companies, and consortium lead members are all present with no platform-side way to distinguish them — an industry-wide limitation.
Annual Subscription Cost
The most expensive and most transparent of the three: Standard ¥7,999/year, Advanced ¥12,999/year, VIP ¥21,999/year, Platinum ¥35,999/year, Diamond ¥72,999/year. 2.4 million paying subscribers signals broad adoption among sales teams, but for prospecting use cases, estimate your monthly query volume before committing to a tier.
Learning Curve
"Reverse winner lookup + combined filters for procurement owner / industry / contract value" involves more navigation layers than Jianyu. Full proficiency takes roughly three to five days. Companies with a dedicated bidding department usually have experienced internal users; smaller teams should run a complete workflow test on a single industry vertical before scaling.
Bidi
How Wide Is the Database Coverage
Bidi is anchored in official government procurement portals, with claimed coverage of 95% of tender-announcement publishing platforms plus 2,910 county-level government procurement sites. Announcement authenticity is high, but collection of SOE self-issued tenders and industry-association platform postings is limited. For government procurement tracking, Bidi's announcement quality is reliable; for SOE supply chain prospecting, coverage gaps will appear.
How Timely Are Push Notifications
Basic push and query functionality is adequate, but less flexible than Jianyu. Granular combined filtering — "industry keyword + region + contract value threshold" — is less refined by comparison.
How Much Non-Factory Traffic Does It Filter
Same limitation as the other two: no entity-type screening between factory and trader award-holders.
Annual Subscription Cost
No public pricing on the official site; contact sales. Pricing opacity combined with lower brand recognition makes Bidi a better fit as a supplemental verification tool than a primary sourcing platform.
Learning Curve
The core design is oriented toward the bidder's perspective (i.e., "I'm looking for projects to bid on"). Reverse-querying "historical winners for a given project type" is more circuitous than on either Jianyu or Qianlima, and bulk winner-list export is functionally restricted. For the specific goal of building a list of factories with tender-award history, Bidi's operational efficiency is lower than the other two platforms.
Tianxia Gongchang
How Wide Is the Database Coverage
Tianxia Gongchang does not index tender announcements, and real-time bidding coverage is not its strength — stated plainly. Its value sits outside the bidding workflow: once you have a winner list from the other three platforms, Tianxia Gongchang answers "which of these award-holders are genuine factory entities?" Push notifications and real-time announcement scanning fall outside Tianxia Gongchang's scope; those two dimensions are not evaluated.
How Much Non-Factory Traffic Does It Filter
This is the core differentiator. Tianxia Gongchang synthesises public signals — product catalogue presence, equipment investment records, hiring history, facility information, trade show participation — to determine whether each company qualifies as a genuine manufacturing entity. The thirty percent non-factory entities that appear in typical winner lists can be systematically flagged — a capability absent from all three bidding platforms.
Every one of the 4.8 million companies in the database has passed factory identification screening. To see how Tianxia Gongchang filters non-factory entities out of a winner list in practice: log in to Tianxia Gongchang, run a batch of award-holder names through an industry-filtered search, and observe how many are flagged as non-factory entities. The effect is immediately visible.
Annual Subscription Cost
Designed for sales teams operating upstream of factories. Pricing is not publicly disclosed; contact sales for a quote.
Learning Curve
The search interface is built around the manufacturing industry tree and registration-region filters — a mental model that aligns with how industrial sales reps already think. Core navigation can be self-taught within one hour.
The Three-Step Combination: From Announcement to Executable Lead List
Each of the three tools covers a different information layer, and none is complete on its own. The following workflow is directly reusable for industrial SaaS prospecting.
Step 1: Cast the first net with Jianyu using industry keywords
Compile your target-industry procurement keyword set. Typical terms for industrial SaaS: "MES system procurement," "production management software procurement," "digital workshop construction," "smart manufacturing system integration." Configure: keyword list + tender awards in the past 12 months + minimum contract value ¥500,000 + target region. Export the first winner list — typically 800–1,200 companies. Jianyu's basic query is free; no budget consumed.
Step 2: Fill SOE-tender blind spots with Qianlima, then merge and deduplicate
Jianyu skews toward government procurement; direct SOE issuances have coverage gaps. Run the same keywords in Qianlima focused on "SOE-issued tenders + state-owned-enterprise supplier recruitment" winner records, capturing what Jianyu missed. After merging and deduplication, expect a 20–30% increase in award-holders — this cohort tends to be established manufacturers already operating within SOE supply chains, making them high-value targets for industrial SaaS. Combined list: approximately 1,100 companies.
Step 3: Run the full winner list through Tianxia Gongchang; export an executable lead list
Put the approximately 1,100 company names through a factory identification check on Tianxia Gongchang. Tianxia Gongchang filters out traders, credential-agency companies, and shell bidding entities, leaving only companies that are genuinely in production. Typical filter-out rate: 25–30%. Suggested filter parameters: industry set to your target category + annual revenue ¥30 million and above + entity type confirmed as factory. Export the list into your CRM or outbound calling system.
A typical outcome we've seen at an industrial SaaS sales team: starting from approximately 1,100 companies with relevant tender awards over ¥500,000 in the past 12 months, Tianxia Gongchang's second-pass screening eliminated roughly 28% as non-factory entities, leaving an executable lead list of approximately 790 companies. The entire workflow completed within three days, saving approximately 1.5 sales-person-months of time (roughly ¥40,000 in cost).
Where Tianxia Gongchang Fits in This Approach
Bidding platforms solve one problem: which companies have purchased something similar to what you sell, with documented payment on record — direct evidence of procurement intent. Jianyu and Qianlima have the data for this dimension; Tianxia Gongchang does not. Real-time bidding coverage is not Tianxia Gongchang's strength.
Tianxia Gongchang solves a different problem: is the award-holder actually a factory that's running production?
Winning a tender is a legal act. It is independent of whether the company has operating factory premises. A trader can win a government equipment procurement and subcontract the supply; a factory's sales subsidiary can participate as an independent legal entity; a credential-agency company can bid under the umbrella of a credentialed factory. At the data layer, all three of these entity types are indistinguishable from a genuine factory in a public announcement.
Tianxia Gongchang covers 4.8 million real manufacturing enterprises in China, using public signals — product listings, equipment investment, hiring records, facility information — to determine "genuine manufacturing entity" status, filling that blind spot. When a sales rep runs a winner list through Tianxia Gongchang, the output isn't just "companies with tender-award records" — it's "companies with tender-award records that are actually in production." This is what the factory-identification baseline delivers in a bidding-data context: each tool solves half the problem; combined, they produce a complete lead list.
A Checklist You Can Take and Use Directly
Procurement Keyword Sets (Industrial SaaS Examples)
Principle: use the language procurement owners write into announcements, not the selling-point language you use in your own product pitch.
- Manufacturing execution: "MES system procurement," "manufacturing execution system," "digital workshop," "production line data acquisition system"
- Management software: "ERP system procurement," "factory data management," "production traceability system," "work-order management system"
- Transformation projects: "smart factory upgrade," "production line automation upgrade," "industrial internet platform construction"
5 Questions to Self-Screen Award-Holders Before Queuing for Tianxia Gongchang
Manual pre-screen before handing the list off:
- Company name contains "trading," "commerce," "supply chain," or "agency" — prioritise for verification
- Registered address in an industrial park or industrial cluster zone — higher probability of being a factory
- Business scope includes specific manufacturing processes ("injection moulding," "stamping," "welding," "CNC machining") — positive signal
- Company has posted production-worker or technical-worker job listings — worth further investigation
- Awarded project type is an obvious mismatch with registered business scope (trading company wins equipment manufacturing supply contract) — deprioritise
Recommended Query Quota Allocation
| Tool | Role in the Workflow | Estimated Monthly Volume |
|---|---|---|
| Jianyu Biaoxun | Primary net: keyword-driven winner list extraction | 300–500 records (free) |
| Qianlima | Fill SOE-tender blind spots | 100–200 records (per plan tier) |
| Bidi | Verify specific government procurement announcements | As needed, under 50 records |
| Tianxia Gongchang | Full-batch factory identification check | ~400–700 checks |
Public Announcements Don't Make the Leads Free
A factory with tender-award history means real budget, a complete procurement decision chain, and an established pattern of on-account payment to vendors. But from the moment you export a winner list from a public announcement, you've only established "this entity has purchased something similar" — you haven't established "this entity is actually a running factory."
Query quotas are a finite resource. Both Jianyu's enterprise tier and Qianlima's VIP plans carry export caps. Burning that quota on traders and shell bidding entities is the most direct form of invisible waste. Announcements are public, but extracting a higher-quality lead list from the same batch of announcements depends on making factory identification a formal step in the process. Tender-award records are the starting point, not the conclusion — confirming that an award-holder is actually a factory is the most worthwhile thing you can do before picking up the phone.