When a company loses a public tender, the first explanation it tells itself is almost always the same: “they gave it to the cheaper bidder” or “it was rigged”. But the analysis of real evaluation reports paints a very different picture. Read word by word — by a specialised language model that reads whole Bulgarian protocols — hundreds of committee decisions show that the most common reason a serious firm drops out is not price, but a document.
This piece is built on the full participation data of nine active firms on the Electronic Public Procurement Platform — over 7,100 bids across about three years — plus the automated reading of hundreds of their evaluation reports. No names are mentioned; what matters are the patterns by sector.
First — about win rates, honestly
It is easy to be misled by a single figure. If you take only tenders with a clear outcome, some of these firms look phenomenal — two of them, both in medical and laboratory equipment, win about 70% of decided tenders. But that is a misleading number.
The difference comes from the fact that for many tenders there is simply no published data on who won. So we keep both denominators. At one pole — specialists with a double-digit win rate even against all bids. At the other — a construction firm that submitted 2,644 bids (more than one a day) and won just 1%.
The two universes of losing
The strongest takeaway from reading the reports is that the reasons for losing cluster by sector — surprisingly cleanly. There are two universes, and a firm must know which one it lives in.
Universe one — documents. At construction and specialised firms, the absolutely dominant cause of dropping out is irregular or missing documentation. At one construction firm this type of cause is behind nearly six in ten analysed losses. The wording from the reports is almost painfully petty: unfilled national databases in the ESPD; missing certificates of experience for the site manager and the quality controller; arithmetic errors in the bills of quantities; a missing specification translation; a delivery deadline given in working instead of the required calendar days. Not one cause is about money — each is about compliance.
Universe two — price. At goods and energy traders the picture flips. Here the dominant cause is an “abnormally low price” — the firm bid so low that it falls into the mandatory check under Art. 72. At one trader this motive is behind nearly half of losses. These are markets — fuel, consumables, food vouchers — where the subject is standardised and the battle is entirely about price. In one report the three bidders offered an identical price and the winner was decided by lot.
The anatomy of a disqualification
Combining all causes, one distinction stands out: the difference between losing and being excluded. To lose means your bid was scored and ranked lower — a fair loss. To be excluded means it never reached evaluation.
At one construction firm nearly four in ten failures are because something in the file was wrong. At a goods trader — just 2%: it almost never drops out formally, always reaches evaluation, and loses (when it loses) fairly, on price. The difference reflects documentation complexity by sector.
Portraits by sector
The volume builder. Bids on everything — roads, roofs, renovations. Thousands of bids, a single-digit win rate. Rarely loses on price and almost always on a document. Its paradox: it works hard in the wrong direction — more bids instead of more compliant ones.
The specialist in a regulated niche. Medical and laboratory equipment. Faces the strongest competition — eight rivals per tender on average — and still wins the majority of decided ones. When it loses, the cause is a technical subtlety: invalid authorisation letters, a catalogue “not in the required manner”.
The IT supplier. A balanced profile. Its characteristic losses are concrete: “no manufacturer, brand and model”; “no specification translation”.
The standard-goods trader. Fuel, toners. Faces competition in practically every tender. Loses fairly, on price, and often falls into the abnormally-low-price check.
The quiet tenders and the power of competition
The data also allows a look at something rarely discussed — how “quiet” some tenders are.
A single-bidder tender is not necessarily a violation, but a high share deserves attention: either the subject is defined too narrowly, or the others decided it was not worth it. The reverse is instructive: the strongest firm wins the majority of tenders despite facing eight rivals on average. That is real market power — a win in a full room is worth more than a win in an empty one.
A surprise: winner data lies more often than we expected
The reading revealed something methodologically important. When we compared the officially published contract data with what the reports say, mass discrepancies appeared. In dozens of cases the official record listed the firm as “did not win”, while the report clearly ranked it first.
The explanation is almost always the same: lots. A tender is often split into several lots; a firm can win one and lose the others. Public records easily “stick” the wrong winner to the whole tender. The takeaway: don’t trust aggregated records blindly — the truth is in the report. And for business: your own record of won and lost tenders in public registers may be less accurate than you assume.
What actually distinguishes the winners
First, compliance above all. In a world where most losses come from documents, not price, the winner is the one whose bid does not drop out on a technicality. An internal “checklist” review before every submission would save a huge share of losses.
Second, specialisation over volume. The high-win-rate firms don’t submit the most bids — they submit the most targeted. More participations don’t mean more wins; they mean more cost.
Third, knowing your own market. On the price market they enter with a real, defensible price. On the document market — with a flawless file. And they don’t waste resources in others’ niches.
The lessons for business
- The document decides more often than price. Before optimising your price to the last cent, make sure your file won’t exclude you before evaluation.
- Know whether you are on a “price” or a “document” market. Standard goods → a price battle (with a ready justification). Construction/complex services → a document battle.
- Count wins, not bids. Two hundred flawless bids beat two thousand rushed ones.
- Don’t be fooled by “first place” or your own record. The opening order barely predicts the final result, and public records err — especially for multi-lot tenders.
If the most common cause of loss were price or connections, the small firm would be powerless. But when the cause is a document, it is entirely in its own hands. More often than anything else, the winner is not the cheapest or the best-connected, but the most compliant.
This is exactly what underpins our competitor & history analysis and our requirements ↔ documents check.
— Analysis of open EOP data and automated reading of evaluation reports.
