Firm Awards Intelligence · Phase 1 · World Bank

Who actually wins the contracts.

Every Contract Award notice published by the World Bank for ASEAN-11 project countries — winning firm, losing bidders, all three prices (bid open · evaluated · signed), procurement method, evaluation scores where published, and the underlying project. Backed by the WB procurement API, refreshed monthly. Built so the data team can benchmark firms, learn from bid patterns, and see who is showing up in our markets.

Source: World Bank procnotices API Coverage: ASEAN-11 · last 3 years Updated: — Next expansion: ADB + 10-year backfill (roadmap)

Coverage snapshot

Numbers reflect all sources loaded in your browser. Signed values sum in USD equivalent for cross-currency comparison.

Module 2 · New

Who bids with whom

This table names the winner. It rarely tells you the winner was three firms in a joint venture — or that the same three have now won together five times. The JV network map rebuilds those alliances from the award data behind this page: every consortium, who led it, who came along, and how the pairings shift year to year. Every awarded firm is also run against the World Bank debarment list, which carries the cross-debarments from ADB, AfDB, EBRD and IDB.

Joint ventures 401 Firms mapped 770 Partnerships 1,091 Debarment flags 5
Open the JV network →
Module 3 · New

Where the winning bid actually sat

Every bidder on every award — winners and losers, the price each put in, and on quality-and-cost tenders the technical and financial scores the evaluation published. Filter down to the tenders that look like yours and read the distribution: how far above the cheapest bid winners really priced, how wide the field spread, and how often lowest price simply took it.

Tenders 5,788 Contested 1,560 Lowest price won 76% Scored 673
Open Bid Analytics →
Module 4 · New

Follow one firm all the way through

Pick any firm and see the whole file: which ministries it works for, who it forms joint ventures with and who leads them, which firms it keeps meeting in the same competitions and who wins those, and what it has won or lost year by year. Where a ministry has published its evaluation sheet you also get the score the firm got when it lost.

Firms 8,334 Ministries 140 Awards traced to a client 96%
Open Firm Tracker →

Top firms

Firm-level rankings — deduped by normalised name. Search a firm to find it. Click any name to see all its contracts in the table below.

Most wins

Ranked by number of Contract Award notices held.

    Most contract value

    Sum of signed contract price, USD equivalent.

      Most active bidders

      Total distinct projects touched (won + evaluated).

        Full awards table

        Every award-bidder row in the current window. Click a column header to sort. Click a project name to open the WB source notice.

        — rows shown
        Date Firm Role Country Project · Scope Method Signed price

        Methodology & roadmap

        Where the data comes from, what it does and doesn't cover, and what's next.

        Sources · what's live and what's ready

        World Bank — fully automated. Procurement Notices API (search.worldbank.org/api/v2/procnotices) filtered to notice_type=Contract Award across ASEAN-11. Each WB notice ships a semi-structured HTML blob in the notice_text field with winning firm(s), any evaluated (losing) bidders, opening / evaluated / signed prices, contract duration, and — for consulting contracts (QCBS) — full technical + financial + combined evaluation scores with rank. Parser at scripts/scrape-awards/lib/parse-wb-notice.js.

        ADB · AIIB · IFAD · IsDB · JICA · KOICA · AFD · GCF · USAID · UN Global Marketplace — ready to ingest via quarterly CSV drops. None of these publishers exposes a Node-cron-scrapeable API (ADB uses Oracle AF sessions, IFAD publishes only PDFs, UNGM is login-walled, adb.org/data.adb.org sit behind Cloudflare). All of them DO publish their own official quarterly Awarded-Contracts CSV/Excel exports via their public UI. Drop those exports into data/awards-manual/ — the pipeline auto-detects the source from the filename prefix (adb-, aiib-, ifad-, isdb-, jica-, koica-, afd-, gcf-, usaid-, ungm-) and routes each through its own column mapper in scripts/scrape-awards/lib/csv-import.js. Each source gets its own colour-coded badge on the page. Full step-by-step download instructions per publisher are in data/awards-manual/README.md.

        Refresh: WB — node scripts/scrape-awards/index.js --backfill anytime. Manual sources — drop a fresh CSV in data/awards-manual/ and run node scripts/scrape-awards/index.js --skip-wb.

        Roadmap

        Next: Expand WB backfill from 3 years → 10 years so firm win-rate trends stabilise.

        Monthly GitHub Actions cron that runs the WB fetch + republishes automatically (manual CSV drops stay quarterly, human-in-the-loop).

        Firm profile pages — click a firm to see all its awards, average bid/signed spread, competitors it bids against, win-rate by procurement method.

        Cross-source firm deduplication (WB "PRACTICA" vs ADB "Practica Ltd") + price normalisation using contemporary FX rates.

        When ADB / IFAD / UN eventually publish a machine-readable feed we swap the manual-CSV importer for a live scraper — the on-page schema stays identical, so the transition is invisible to users.

        Data limits (be aware)

        Losing bidders not always public. Many WB notices publish the winner only; evaluated-bidder blocks appear mainly on Quality-and-Cost-Based Selection (QCBS) and higher-value ICB contracts.

        Consortium awards attribute the signed price to one consortium member, so a firm's total here counts the contract once against whichever member signed it. The other members of each joint venture — and who led it — are mapped on JV Network Intelligence.

        USD equivalents use static exchange rates (rough long-run averages). Fine for ranking; not for accounting.

        WB "Contract Award" notices are not the same as procurement volume — a project might have 100 awards or 5, depending on how it's chunked. Firm rankings are best read as "who shows up," not "market share."