‘Clear explanation’, ‘necessary guarantees’, ‘in line with the market’: GO! finds four abnormal prices and explains in three phrases why they are normal — the Council of State suspends
For its Plan Management framework agreement, GO! flagged all four offers as abnormal, allowed the lowest bidder to double its amount from 153,540 to 317,340 euros after a ‘material error’, and accepted the price justifications of the two highest-ranked bidders on the finding that they had given ‘a clear explanation’ — reasoning the Council of State called a style formula, and enough to suspend the award under extreme urgency at the request of the third-ranked bidder.
What happened?
GO! Onderwijs van de Vlaamse Gemeenschap, the Flemish Community education network, launched an open procedure for a services contract entitled ‘Framework agreement – Plan Management’: keeping up to date the floor plans and associated data in the Facility Management Information System (FMIS) and the point clouds and photographic material in the Orbit viewer. GO! acted as a central purchasing body within the meaning of article 2, 4°, of the Public Procurement Act, for all its school groups and for itself. The specifications, reference GO/ROV/001PLANB/2026/7143/PB, concerned a price-list contract with a maximum value of 1,300,000 euros including VAT over the whole term, with the express warning that this ceiling had not been calculated on the estimated quantities in the inventory and was therefore no indication of the tender amount. The contract was to be awarded on the best price-quality ratio: price 60 points under the formula Pmin/P, quality of the approach plan 10 points, and a proof of concept 30 points — an on-site survey and drawing of a recently modified building of about 1,000 square metres, scanned in colour, assessed against standard EN 15221-6 and to be carried out in half a day. By the deadline of 27 May 2026 at 11.00 a.m., four offers had been submitted. Excluding VAT, company G. offered 399,060.00 euros, company V. 791,200.00 euros, the applicant O. 1,041,700.00 euros and L.V.E. 153,540.00 euros. GO! compared each price with the average of 596,375.00 euros and applied a maximum deviation of fifteen per cent. All four offers fell outside that band and were recorded as ‘abnormal’ in the verification report: O. was 74.67 per cent above the average, V. 32.67 per cent above, G. 33.09 per cent below and L.V.E. 74.25 per cent below. On 2 June 2026 GO! asked L.V.E. for a justification under article 36 of the Royal Decree of 18 April 2017, focusing on the unit prices for data collection with and without clearing; the same day it asked G. for further information on its price structure, to which G. replied on 8 June. L.V.E. returned the question: on 3 June it first sought clarification on the scope of the items, because it had understood the data-collection items to cover only the on-site survey and the processing of the measurement data, not the full preparation of the DXF plan. On 5 June GO! confirmed that those items did cover the full combined performance and that the separate drawing item applies only where no new survey is needed. On 10 June L.V.E. filed its price justification, explaining that it had not priced the DXF work into those unit prices. GO! treated this as a material error and corrected the offer: since the drawing prices were already known at the opening, a simple addition sufficed. The corrected amount became 317,340.00 euros excluding VAT, or 383,981.40 euros including VAT — more than double the amount opened. After that correction the average moved to 637,325.00 euros, with L.V.E. still 50.21 per cent and G. 37.39 per cent below it. The verification report of 9 July 2026 concluded: ‘Both bidder [L.V.E.] and bidder [G.] provided a clear explanation in good time. This also shows that they can offer the necessary guarantees for performance of the contract. The latter is moreover addressed under the proof of concept. The above price analysis confirms that the offers are in line with the market and contain no abnormal prices.’ All four offers were found regular. The ranking became L.V.E. 84 points, G. 63.71, O. 50.28 and V. 28.07. On 15 July 2026 the managing director of the Community Education network endorsed that proposal and awarded the contract to L.V.E.; O. was notified on 16 July and on 28 July received a partial answer to its request for further information, subject to confidentiality. On 30 July 2026 O. sought suspension under extreme urgency, raising one plea in two limbs: that the price adjustment for L.V.E. was not a permissible correction of a material error but a prohibited modification of the offer after opening, and that the acceptance of the price justifications of both the first and the second-ranked bidders had been neither carefully examined nor adequately reasoned. GO! and L.V.E. objected that O., ranked third with an offer of 1,260,457.00 euros including VAT against 383,981.40 euros, had no interest. The Council rejected that objection: the interest in regaining at least a chance suffices, and because the criticism also touched the second-ranked bidder, a fresh price analysis could place O. first. The Council first made a procedural observation. Since 1 January 2025, article 7, § 1, of the Royal Decree of 19 November 2024 requires the defending party to include in its observations a summary of its own arguments whenever answering the pleas calls for a fuller exposition, and that obligation applies in extreme-urgency proceedings too. GO! had summarised O.’s limbs — which O. had already summarised itself — but not its own arguments, spread over several pages. The Council noted that it is not for it to do in that party’s place what that party must do, and that GO! must accept that its arguments are understood as the Council went on to understand them. On the substance the Council sided with GO! on several points. O.’s treatment of the maximum value of 1,300,000 euros as an estimate was prima facie unconvincing: the specifications expressly state that this ceiling is no indication of the tender amount. GO! was also entitled, within its margin of assessment, to use a fifteen per cent deviation from the average as a detection criterion, even though article 36, § 4, did not apply here. Nor is it required that every element of a price justification be unique to the bidder concerned: automation, work organisation or geographical spread may, in combination, produce a relevant cost saving. And GO! could confine the closer price analysis at this stage to the first and second-ranked bidders; as a legality judge the Council does not substitute its own price assessment for that of the authority. The Council did observe that GO! had referred, in its own correspondence, to ‘the estimate we put forward for the contract’, which prima facie suggests that, apart from the maximum amount in the specifications, an estimate did exist against which the offers appear to have been tested. The core question remained whether the price justifications removed the established appearance of abnormality and whether their acceptance had been carefully examined and adequately reasoned. That is where it failed. The words ‘clear explanation’, ‘necessary guarantees’ and ‘in line with the market’ contained, in the Council’s view, the conclusion of the assessment rather than the reasons leading to it. The report does not state which parts of the justifications were decisive, nor how they explain the deviations of 50.21 and 37.39 per cent. For the successful bidder the report discusses at length why the tender amount was adjusted, but not why the corrected amount, still half below the average, was economically feasible; on the further price information of 23 June 2026 it contains no concrete assessment. For the second-ranked bidder virtually any individual discussion is missing. The reference to the proof of concept does not cure this: a practical test may show that a bidder grasps the technical scope and can deliver a result, but prima facie does not show that the whole contract is economically viable at the unit prices offered over the term of the framework agreement; it can at most be a supporting element and does not replace a concrete analysis of staff costs, productivity, overhead, travel and margins. Grounds raised by GO! afterwards in its observations could not remedy the defect either: they appear in no document of the administrative file and GO! did not show that they played any role when the decision was taken. Because the defect affects both the first and the second-ranked offer and a fresh price analysis could affect O.’s ranking, O. had an interest in its criticism. The second limb was found serious. Councillor of State Kaat Leus, acting president of the XIIth vacation chamber, admitted L.V.E.’s intervention on 24 August 2026 and ordered the suspension of execution of the award decision of 15 July 2026 under extreme urgency. First auditor Alexander Van Steenberge had given a concurring opinion. On the first limb — whether doubling the tender amount from 153,540 to 317,340 euros excluding VAT was still a permissible correction of a material error — the Council did not rule in this judgment.
Why does this matter?
This judgment says little about what an abnormally low price is, and a great deal about what an authority must write down once it suspects one. GO! did not do little: it calculated an average, applied a fifteen per cent threshold, questioned the two lowest bidders, received answers and placed them on file. The Council expressly approves each of those steps. What was missing was the last one: explaining why those answers remove the appearance it had itself established. A detection threshold you apply yourself creates an obligation: whoever selects a price as abnormal on the basis of their own method must then actually follow it up carefully. Three reassuring phrases are then not reasoning but a conclusion without it. Two consequences matter in practice. First, confidentiality is no free pass. The Council expressly accepts that an authority may stay in general terms to protect business data — but not so briefly that it can no longer be checked why the offer was found regular after the price analysis. And repairing it in the observations does not work: grounds absent from the administrative file do not count, otherwise the duty to state reasons is hollowed out entirely. Second, a proof of concept proves capability, not profitability. That a bidder surveys and draws correctly in half a day says nothing about whether it can deliver at those unit prices for four years. Using the test as an answer to a price objection conflates two different examinations. The judgment is also useful for bidders ranked well down the list. O. stood third with a tender amount more than three times the winner’s and looked hopeless at first sight. Its interest was nevertheless accepted, precisely because its plea touched not only the winner but also the runner-up: if both offers are rejected after a new price analysis, it moves to first place. Here the drafting of the application determined admissibility. Finally there is a procedural novelty. Since 1 January 2025 the observations must contain a summary of the party’s own arguments, in extreme-urgency proceedings too. Failing to do so risks the Council reading your arguments as it reconstructs them — a disadvantage you cannot repair in an accelerated procedure.
The lesson
As a contracting authority the lesson is compact: what you establish in the price analysis, you must also finish in the award decision. If your own method selects a price as apparently abnormal, write down which concrete elements of the justification you found decisive, how they explain the established percentage deviation, and why the price is economically feasible over the whole term. Do so per bidder: one joint formula for two different files is not reasoning. If the justification is confidential, stay general in wording but concrete in structure, so that a reader can verify what you examined. Keep those grounds in the administrative file, because what you raise only in your observations does not count. And do not use a proof of concept as a price argument — that test shows capability, not profitability. As an unsuccessful bidder, read the verification report for its verbs. Does it state what the authority established, or only what it concluded? Sentences such as ‘the offers are in line with the market’ without identifiable analysis are your point of attack. Look above all for self-contradiction: an authority that first lists four prices as abnormal and then declares them normal in a single sentence has a gap in its file. And if you are ranked third rather than second, aim your plea expressly at every bidder ahead of you — here that is exactly what defeated the objection of lack of interest.
Ask yourself
Does your award decision explain, bidder by bidder, which concrete elements of the price justification you accepted and how they explain the established percentage deviation, or does it stop at a conclusion? Are those grounds in the administrative file, or would you have to raise them for the first time in your observations? Are you using a proof of concept, a reference or a quality score anywhere as an answer to a price objection, even though these say nothing about the economic viability of the unit prices? Have you checked whether your file contains an estimate you refer to in correspondence, and whether that estimate is consistent with what the specifications say about the maximum value? And as a bidder: does your plea touch every bidder ranked ahead of you, so that a fresh analysis can actually place you first?
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 →