Two families of criteria
Criteria fall into two families with different procedural treatment. Automatic criteria are calculated and judgement based ones are argued.
Formula-based (automatic) criteria are scored by applying a mathematical operation to an offered figure: the price, an extended warranty period, a reduced execution period, quantifiable commitments. Their scoring admits no discretion.
Judgement-based criteria require a technical assessment: the quality of the construction method statement, the works programme, quality control or environmental management measures. They are scored before the automatic criteria are known, precisely so the judgement is not contaminated by the price, and when they outweigh the automatic ones their evaluation falls to a committee of experts.
In the simplified open procedure, judgement-based criteria are capped at 25 per cent of the score in works, which turns most mid-sized tenders into essentially automatic competitions, as explained in the article on the open and simplified open procedures. Strategy then shifts towards price and the automatic criteria.
The price formula: where tenders are decided
Among the automatic criteria, the price scoring formula is the most decisive piece and the least read. Two tenders with the same weighting (say, 60 points for price) can produce opposite outcomes depending on the chosen formula:
| Formula type | Behaviour | Strategic consequence |
|---|---|---|
| Purely proportional to the discount | Each extra point of discount adds points linearly | Rewards aggressiveness: the maximum viable discount is the optimum |
| Proportional to the offered price | Small point differences between close bids | Price barely discriminates; the technical part decides |
| With a saturation threshold | Beyond a certain discount no more points are earned | Discounting past the threshold gives away margin for no return |
| Referenced to the average discount | Points depend on where the average of all bids lands | Forces an estimate of competitors' behaviour |
Average-referenced formulas bring game theory into the tender: the optimal bid depends on what everyone else does, and analysing the authority's previous tenders (public on the contracting profile) becomes first-order competitive intelligence. Review tribunal case law also requires formulas to distribute the available points effectively; formulas that in practice hand almost all points to any symbolic discount have been annulled repeatedly.
Improvements, deadlines and other automatic criteria
Improvements (additional work at no cost) are only admissible if the documents define them precisely: what is admitted, on which elements and how they are scored. Generic improvements ("proposed improvements will be evaluated") are unlawful because they leave the award to the committee's discretion. For bidders, defined improvements are in reality hidden cost: they must enter the cost study as one more item before being committed.
A reduced execution period as a criterion deserves double caution: the points are earned in the bid, but the shortened period becomes a contractual obligation subject to penalties. Offering an impossible schedule to scrape points is financing the tender with future penalties, as analysed in the article on deadlines, penalties and extensions.
Reading the weighting as strategy
The practical analysis of the criteria comes down to three questions. First: where are the points? A 70/30 price-weighted tender is played in the cost study; a 50/50 one, in the technical submission as much as in the discount. Second: how does the formula behave? Simulating the score under three discount scenarios (conservative, medium, aggressive) against hypothetical competitors reveals whether the marginal effort pays. Third: which criteria are cost? Improvements, warranty extensions and schedule reductions have a price; adding them to the cost study avoids winning the tender while losing the margin.
One lower bound remains: the discount resulting from that strategy must clear the abnormality filter of article 149, whose workings are developed in the article on abnormally low tenders. Crossing the threshold triggers a duty to justify, and failing to justify excludes.
Worked example: the same tender under two formulas
Three companies bid for a project with 60 points assigned to price. Bids: A discounts 10 per cent, B 15, C 20. Compare the points distribution under two frequent formulas:
| Bid | Formula proportional to the maximum discount (60 × discount/max discount) | Formula with a saturation threshold at 15% |
|---|---|---|
| A (10% discount) | 30.0 points | 40.0 points |
| B (15% discount) | 45.0 points | 60.0 points |
| C (20% discount) | 60.0 points | 60.0 points |
Under the proportional formula, C beats B by 15 points: aggressiveness pays, and the tender is decided in the cost study. Under the saturation threshold, B and C tie at 60 points, so C's additional 5 points of discount are margin given away with no return at all: the tender will be decided on the judgement-based criteria, and the optimal strategy was to bid exactly at the threshold.
The operational conclusion: before deciding the discount, simulate the formula with scenarios rather than skim it. Ten minutes of spreadsheet work separate the calibrated bid from the one that burns margin for nothing.
Social and environmental criteria: points that are also cost
The LCSP pushed qualitative social and environmental criteria, and in works they are now standard scenery: reinforced waste management plans, recycled material percentages, low-emission machinery, hiring of long-term unemployed people, on-site training. They are scored as automatic criteria (quantifiable commitments) or within the judgement-based part (quality of the proposed measures).
The correct treatment in the bid is the same as for improvements: every scoreable commitment is a cost item to be quantified before offering it, because after signature it becomes a verifiable and penalisable execution condition. The symmetrical error also exists: systematically renouncing those points hands an advantage to competitors who have industrialised their answers (a standard waste plan, stable agreements with sheltered employment centres) and offer them at a marginal cost that decreases tender after tender.
Strategic reading of the documents therefore includes an inventory of the social and environmental points reachable with the company's current structure and of those that would require investment: the former are always offered; the latter, only if the formula analysis shows those points decide the tender. The inventory is redone for each tender, because the criteria change.
Note: the legal references cited reflect Spanish legislation in force in 2026 (Ley 9/2017). Always check the criteria and formulas in the documents of each specific tender.