A weighted model does not exist to avoid price. It exists because price and specification are one decision, and a bid price means nothing until the offer behind it has been read.
This is a fault-to-fix guide for the evaluation stage. It sets out what a price-only award produces, which criteria actually predict the cost of a blade, how to weight technical and commercial factors, and how to document a decision that will be reviewed by people who were not in the room.
Evaluating plow blade tenders: why lowest price fails
The lowest bid is cheapest only when offers match.
A plow blade tender that is awarded on price alone is comparing quotations rather than outcomes. The consequence is not that the lowest bidder is dishonest; it is that where the specification leaves room, the lowest bid tends to be the lightest compliant reading of it.
The difference then appears as a shorter service interval, more change-outs and more downtime, none of which is visible on the invoice that decided the award.
The second failure is subtler and more common: the price is comparable only if the offers are. Where one bidder included freight and another did not, or one quoted per metre and another per edge, the arithmetic of the comparison is being done on unequal figures. Normalising the offers is therefore the step that has to happen before any scoring, and it is set out on the supplier quote comparison page.
What a price-only award produces
Three outcomes follow from a price-only award, and each has a predictable shape. The winner is the offer that made the fewest commitments: the minimum tolerance, the least documentation, the shortest stated interval. The evaluation produces no record of why the technical differences did not matter, so the decision cannot be explained if it is challenged. And the following season’s tender inherits the problem, because nothing in the process produced evidence about what the delivered edges actually did.
The third is the one that compounds. A fleet that awards on price and never records the interval it receives has no basis for a different award next year, so the same decision is made in the same way with the same result.
Choosing criteria that predict cost
Criteria should be chosen for their ability to predict the cost of the outcome rather than for their tidiness. Four do that work: the service interval the product can deliver on the fleet’s routes, the labour each change-out consumes, the documentation that makes a failure resolvable, and the price against a defined specification.
Two further criteria are worth including where the fleet has a specific exposure. The lead time the supplier can hold during the season, since an offer that arrives late carries its own cost. And the warranty remedy, valued as something other than a duration, because a replacement part and a replacement including fitting are different offers.
Weighting technical and commercial factors
Weighting is the part of the model that has to be decided before bids arrive, and the table below sets out the range that usually works.
| Criterion | Typical weight | Why it sits there |
|---|---|---|
| Technical compliance and reported values | Highest | Determines whether an offer can be scored at all |
| Service interval evidence | High | The largest driver of whole-life cost |
| Documentation per lot | Moderate | Determines whether a future claim is resolvable |
| Lead time and delivery terms | Moderate | Sets whether supply arrives when the work happens |
| Price against the defined specification | Moderate to high | Comparable once the rest has been assessed |
The weights are less important than stating them in advance. A model whose weights are set after the bids are opened will be shaped by the bids, and an evaluation whose criteria were agreed with the specification is far easier to defend.
Scoring evidence, not promises
The difference between evidence and a promise is that evidence can be checked. A bid that reports material values with the test methods behind them can be compared with another bid doing the same; a bid that states that a product is durable cannot be compared with anything.
Three practices move scoring towards evidence. Requiring values rather than descriptions for every named requirement. Requiring the test method with each value, so the figures are comparable. And scoring what can be checked at delivery, which ties the evaluation to the receiving record rather than to the bid document. Descriptions of the receiving checks themselves are set out on the delivery inspection page.
Handling non-compliant bids
Non-compliance is normal and the model needs a way to handle it rather than a way to avoid it. Compliant bids are scored. Bids with a deviation that does not affect function can be assessed against a defined concession route, and the deviation recorded. Bids that do not meet a mandatory requirement are rejected, with the reason stated.
What weakens an evaluation is the fourth case: treating a non-compliant bid as compliant without saying so. That produces a comparison resting on a specification nobody applied, and it is the most common reason a tender decision is challenged successfully.
Documenting the decision
The record of an evaluation has to be readable by someone who was not present, and four elements make it so: the criteria and weights as published, the scores with the evidence behind each, the treatment of any non-compliant bid, and the normalised prices the commercial scores were calculated from.
Keeping the evidence with the scores is what separates a documented decision from a recorded one. A score without the figure behind it can only be re-argued; a score with the figure can be checked.
Template scorecard
The scorecard below is the one that carries the least ambiguity and the fewest hidden judgements.
| Section | What it records | Source |
|---|---|---|
| Compliance | Met, met with concession, or rejected, per requirement | The schedule of requirements |
| Reported values | The figures supplied and the test methods | The bidder’s technical response |
| Interval evidence | Stated or demonstrated service interval | Bid evidence and the fleet’s own record |
| Documentation offered | The records issued with each lot | A sample record where available |
| Lead time | Stated lead time for the quoted quantity | The bidder’s response |
| Normalised price | Cost per edge and per metre on stated terms | The commercial response, converted |
The first row is deliberately first in the scorecard as well. A bid that has not been assessed for compliance cannot be scored on anything else, and putting compliance at the front prevents an attractive price from being scored before the offer behind it has been read.
What a weighted model cannot fix
Three problems survive a well-built scoring model, and recognising them is part of using one honestly.
- A loose specification, since no weighting can score what was never required.
- A fleet with no interval data of its own, which leaves the largest criterion scored on the bidder’s claim.
- A panel that disagrees about the weights after the bids are open, which turns the model into a negotiation.
The second is the one a fleet can fix before the next tender. A season of recording service intervals on the routes in the tender gives the panel something to score against, and it converts the strongest criterion from a promise into a comparison.
Reviewing the model after a season
The model is verified by the delivery it produced. One comparison is enough: the interval the winning offer actually delivered against the interval that was scored. Where those agree, the evaluation was predicting something real; where they differ, the criterion was scored on a claim rather than on evidence, and the weighting should move next time.
Recording that comparison across seasons is what turns a scoring model into a procurement asset. Cost and performance research is published by the Transportation Research Board, procurement practice by the National Association of Counties and the American Association of State Highway and Transportation Officials, and material test methods by ASTM International and ISO.
SENTHAI manufactures plow edges, wear plates and carbide inserts in Rayong, Thailand, supplies public fleets against specified drawings, and reports the values a weighted evaluation needs with the test methods behind them. Tenders that need the technical response or the documentation set confirmed can be raised through the contact page.
FAQ
Why does lowest price fail in plow blade tenders?
Because the price is for a specification rather than for an outcome. Where the specification leaves room, the lowest bid tends to be the lightest compliant reading of it, and the difference is discovered in change-out frequency and downtime rather than on the invoice.
Which criteria actually predict the cost of a blade?
The service interval the product delivers, the labour each change-out consumes, and the documentation that makes a failure resolvable. Price matters, and it is comparable only once those three have been read against the same specification.
How should evidence be scored rather than promises?
By requiring the bidder to report values with the test methods behind them, and by scoring what can be verified at delivery. A claim about service life that cannot be tied to a batch record or a measured interval is a promise, and it should not carry the same score as evidence.
What should happen to a non-compliant bid?
It should be recorded, assessed against the stated requirements and either rejected or admitted through a defined concession route. What weakens an evaluation is treating a non-compliant bid as compliant without saying so, because the comparison then rests on a specification nobody applied.
Send the criteria you intend to weight and the patterns in the tender. The SENTHAI engineering desk will confirm the values and documentation that can be reported against them.