An ordinal scale with scores to two decimal places: here 'intermediate values allowed' becomes a smokescreen for undisclosed sub-sub-criteria
Zorgbedrijf Antwerpen scores bids for a Menu Management System on an ordinal 50/55/70/85/100% scale with 'intermediate values allowed', but the evaluation report produces scores to the hundredth — which can only be explained by a hidden split into sub-sub-award criteria that appears nowhere in the specifications.
What happened?
Zorgbedrijf Antwerpen launches a competitive procedure with negotiation for a Menu Management System framework agreement: a digital platform for recipes, ordering, food costs, allergens, and links with accounting and POS. Eight candidates apply; five are selected, including the applicant (Q. NV) and the eventual winner (B. BV). After three negotiation rounds and a BAFO, the contracting authority issues an evaluation report on 6 February 2026. The scoring works on an ordinal scale: 100% for 'excellent', 85% for 'very good', 70% for 'good', 55% for 'satisfactory' and 50% or less for 'weak'. The specifications add that 'intermediate values are possible'. The 25 February 2026 award goes to B. BV with 82.77% versus 73.49% for Q. — a 9.28-point gap. The striking part: the awarded scores are fine-grained to the hundredth. For sub-criterion 1.1 (functional fit, out of 30): B. gets 28.40, Q. 16.80, third bidder 23.05. For sub-criterion 1.2 (implementation plan, out of 15): 14.93, 9.61, 11.37. None of that fits a global ordinal scale. The applicant's second plea argues the evaluation method is opaque and inconsistent with the specifications. The authority's defence: the intermediate values reflect 'a nuanced synthesis of input from various expertises'. The Council inspects confidential exhibit 11 — the evaluation sheets — and finds what actually happened: for each (sub-)criterion, the authority converted the non-exhaustive evaluation elements listed in the specifications into (sub-)sub-criteria, each with its own weight, and systematically scored the bids against that breakdown. The two-decimal scores are the arithmetical sum of those sub-scores. But the entire split was never disclosed beforehand. The Council: a contracting authority may allow intermediate values — provided the scores adequately reflect the qualitative differences between bids. Here neither bidders nor a reasonably informed reader of the report could understand how the evaluation committee arrived at its fine-grained scores. The transparency principle and formal duty to state reasons are breached. Even a 9-point gap doesn't rule out that the illegality could affect the ranking. Suspension granted. The Council also rejects the authority's late-filing objection (Labonorm/Neorec line): bidders need not flag illegalities in the specifications during the procurement — they may rely on the presumption of legality and keep their plea in reserve until after the award. The duty of care lies with the authority, not the bidder.
Why does this matter?
Many specifications today say 'intermediate values allowed' on top of an ordinal scale. That's not forbidden as such. But this judgment draws the line: if you end up issuing scores to two decimals, the evaluation report must show how you got there. Otherwise your scoring method is in reality a split into sub-sub-criteria that should have been disclosed up front (cf. CJEU, TNS Dimarso). For bid managers: if an evaluation report contains scores like 28.40 or 16.80 on an ordinal scale, request a detailed calculation before the standstill period ends. For contracting authorities: if your internal scoring template gives each evaluation element its own weight, that breakdown must appear in the specifications — otherwise you breach transparency and motivation, however correct your assessment may be substantively. The judgment also reinforces the Labonorm line: despite pressure from authorities to impose 'Neorec clauses', the Council confirms that bidders carry no 'attention duty' to flag illegalities in the specifications.
The lesson
If you use an ordinal scale and allow 'intermediate values' in the specifications, make sure the final scores in the evaluation report flow logically from the wording of the specifications — not from an internal breakdown that was never disclosed. If your evaluation panel actually works with a spreadsheet in which each non-exhaustive 'evaluation element' carries a fixed number of points, those are (sub-)sub-award criteria and their weight must appear in the specifications.
Ask yourself
If you are writing an evaluation report with scores to two decimals on an ordinal scale: could an outsider reconstruct from your report how you arrived at that exact score, or does the calculation live only in an undisclosed Excel?
About this database
The Council of State (Raad van State / Conseil d'État) is Belgium's supreme administrative court. In disputes over public procurement — from contract awards to tenderer exclusions — the Council of State is the final arbiter. The rulings in this database are summarised by TenderWolf in plain language, with practical lessons for tenderers and contracting authorities. View all rulings →