The formula that scores you

Every bidder reads the award criteria. Price 40%, quality 60%. Fine. They plan the week around it, put the effort where the weight is, and submit.

Almost nobody reads the next paragraph — the one that says how those percentages get turned into points.

That paragraph is the scoring formula, and there is a substantial body of European evidence showing that it can matter more than the weights themselves. It can quietly reweight price. It can make your score depend on companies you've never heard of. In some tenders, it can decide the winner in ways that have nothing to do with the bids.

It is disclosed. It is in the documents. And it is, in our experience, the single most under-read page in public procurement.

Part 1: The weight is not the weight

Start with the thing that sounds impossible.

A tender says price is worth 40%. You assume that means price accounts for 40% of the decision. It's right there in the table.

But a weight is not a score. A weight is an input to a formula, and the formula determines how much your price actually moves your total.

Ballesteros-Pérez and colleagues, reviewing scoring rules used across construction tenders, showed how this breaks. Some economic scoring functions compress the range of price scores. If the formula is shaped such that every credible bidder ends up scoring somewhere between 34 and 40 out of a possible 40, then the price criterion — nominally worth 40% of the total — has a real discriminating power of six points across the whole field.

Price is worth 40% on paper. In practice it separates almost nobody. The tender is decided on quality, whatever the table says.

And the reverse is equally true. A formula that spreads price scores wide — where the gap between the cheapest and the most expensive bid consumes the full range — can make a stated 30% price weight behave, in practice, like 50%. You planned for a quality contest. You entered a price war.

The nominal weights in a tender are not necessarily the effective weights. They are one half of a mechanism, and the other half is the formula. If you only read the first half, you have misread the tender.

Part 2: Your score depends on your competitors

Now the finding that genuinely unsettles people.

Schotanus, Van den Engh, Nijenhuis and Telgen examined 303 Dutch public tenders and simulated the outcomes under the scoring formulas those buyers actually used.

They found that in roughly one in five cases, adding a bid that had no chance of winning changed which of the other bidders won.

A hopeless newcomer submits. They lose, as everyone expected. And the winner — among the serious bidders, who submitted precisely the documents they would have submitted anyway — is now a different company.

The mechanism is called rank reversal, and it is not a mistake. It is a property of relative scoring.

Here is how relative price scoring usually works: the lowest bid receives the maximum price points, and everyone else receives points in proportion to how close they are to that lowest bid. It seems fair. It is certainly common.

But look at what it means. Your price score is not a function of your price. It is a function of your price relative to the lowest one submitted. A bidder who undercuts the field — perhaps recklessly, perhaps because they've made an error, perhaps because they don't intend to win — resets the scale for everybody. Your price, unchanged, now scores differently. And if the gaps between the serious bidders were narrow enough, the ranking above the newcomer can reshuffle.

The study found this was more likely under specific, identifiable conditions:

• Relative rather than absolute scoring
• No minimum quality threshold
• A moderate price weight
• More bidders

Which describes an enormous number of tenders.

The practical consequence is uncomfortable. Under these conditions, a portion of your outcome is determined by arithmetic involving strangers, on bids that were never going to win. There is no proposal quality, no persuasive writing, no beautiful implementation plan that protects you from it.

But you can see it coming. All four conditions are stated in the tender documents.

Part 3: Absolute scoring behaves completely differently

The alternative to relative scoring is absolute scoring: the buyer defines, in advance, a fixed scale. A score of 5 means this. A price below €X earns this many points. The scale doesn't move.

Under absolute scoring, your score is a function of your bid, and nothing else. A competitor's price cannot change your points. Rank reversal is structurally impossible.

Absolute scoring is more work for the buyer — they have to think carefully about the scale in advance, rather than letting the market define it. Which is precisely why it's less common.

So the same bid, submitted to two buyers with identical published weights, can be scored under two mechanisms with entirely different properties. In one, you're competing against a fixed standard. In the other, you're competing against whoever happens to show up.

That is not a detail. That is the game.

What to actually do

None of this requires a mathematics degree. It requires reading three things in the tender documents before you commit a week to the bid:

1. Is the price scoring relative or absolute? Look for the formula. Dutch tenders will often say it explicitly — "de laagste inschrijfprijs ontvangt het maximale aantal punten" (the lowest bid receives the maximum points) is relative scoring, stated plainly. If the tender instead gives you a fixed table of prices and their point values, it's absolute.
2. Is there a minimum quality threshold? A drempelwaarde or minimum quality score. If there is one, bids below it are excluded regardless of price — which sharply reduces rank-reversal risk and means a lowball bid can't drag the field. If there isn't one, price can do more damage.
3. What does the formula do to the range? This is the subtle one. Work out, roughly, what score you'd get and what score a competitor 15% cheaper would get. If the gap is small, price is a weak discriminator regardless of its stated weight, and the tender will be won on quality. If the gap is large, price dominates, whatever the table says.

Then plan the bid around what you found — not around the percentages in the summary table.

Where the weights actually live

There's a practical obstacle to all of this, and it's the reason most bidders skip it: the criteria and formula usually aren't in the notice.

In our own corpus of open EU tenders, fewer than two in five carry award criteria in the published notice at all. And when they do, the data is frequently unusable — duplicated per lot, or filled with placeholders.

Here is one real example. A Dutch tender's TED notice listed every award criterion with a weight of 1. Not 30, not 40 — one, for all of them. Which tells a bidder precisely nothing.

The tender's own document told a different story: 275 points, 275 points, 250 points, and 200 points for its four criteria — plus, of course, the formula that turns those into scores.

The notice is a summary. The document is the contract you are actually being judged under. When they disagree, the document wins, and it is not close.

The uncomfortable conclusion

Public procurement presents itself as a meritocracy: the best bid wins, and the criteria are published so you know what "best" means.

The criteria are published. So is the mechanism that converts them into a decision — and the mechanism has properties that most bidders have never examined. It can reweight price without changing the stated weight. It can make your result depend on rivals. It can, in a meaningful minority of cases, produce a winner that a different formula would not have produced.

None of this is hidden. All of it is in the documents.

The bidders who consistently do well are not, on the available evidence, the ones who write more beautifully. They are the ones who read the mechanism before they decide whether to play — and then, having read it, decide where the week is best spent.

---

FindWell reads the tender's own documents — the Aanbestedingsleidraad and its annexes — and shows you the award criteria and their weights, each quoted with the section it came from. The example above is a verbatim extraction from a real Dutch tender. We work the stages the evidence says decide the outcome: compliance, criterion-fit, and readiness. The writing, the pricing and the submission are your team's, and we don't pretend otherwise.

Sources

• Schotanus, F., Van den Engh, G., Nijenhuis, Y. & Telgen, J. (2021). Supplier selection with rank reversal in public tenders. Journal of Purchasing and Supply Management.
• Ballesteros-Pérez, P., Skitmore, M., Pellicer, E. & González-Cruz, M. (2015). Scoring rules and abnormally low bids criteria in construction tenders: a taxonomic review. Construction Management and Economics.
• González, J. (2019). Weight Values, Scoring Rules and Abnormally Low Tenders Criteria in Multidimensional Procurement: Effects on Price. Hacienda Pública Española.

Procedures, thresholds and standstill periods vary by jurisdiction and are revised periodically. Check the rules in force for your tender.