Scoring Lamp Bids: A Simple Evaluation Matrix for Buying Teams

The familiar problem with comparing lamp quotations is that the numbers are not comparable. One supplier quotes a unit price, another quotes a carton price, a third quotes delivered. Freight is in or out, the bulb is included or not, the warranty runs one year or two. Everyone believes they are comparing like with like, and almost nobody is. A scoring sheet does not remove the judgement, but it does make the assumptions explicit.

Start by levelling the quotes

Before scoring anything, put every quote into the same shape. Convert to a unit price, add the freight allocated per unit, note whether a bulb is included, and record the warranty period. If a supplier will not give you carton dimensions and gross weight, you cannot complete this step, and the comparison should stop there.

Levelling often changes the ranking on its own. A slightly higher unit price in a smaller carton frequently lands cheaper per unit than a low price in an oversized box, because lamp cartons are priced on volume as much as on weight.

The criteria worth scoring

Weight them before you see the numbers, not after. Choosing weights after seeing results is how a scoring sheet quietly becomes a justification for a decision that was already made.

• Landed cost per unit, freight included, at the quantity you will actually order.
• Minimum order, and whether mixed SKUs are permitted against it.
• Dispatch timing measured from order confirmation rather than from payment.
• Documentation: safety listing evidence, marking, country of origin, and how quickly files arrive.
• Serviceability: whether drivers, cords and shades are available as parts, and for how long.
• Finish and batch consistency, which determines whether a reorder matches the installed base.
• Responsiveness during quotation, which is the best available predictor of responsiveness after it.

Scoring what is hard to score

Some criteria are qualitative, and pretending otherwise produces false precision. A five-point scale with a written reason for each score is more useful than a decimal number nobody can defend. The written reason matters more than the number, because it is what survives into the next season when the person who scored it has moved on.

This is the underlying argument in A Scorecard for Lighting Suppliers, which looks at how to keep scoring after the contract is signed rather than only before.

After the scores are in

Look at the spread, not just the winner. A narrow spread between the top two bids suggests the specification is not differentiating anything, which usually means the specification is too loose. A wide spread usually means suppliers interpreted the requirement differently, which means the RFQ needs a second pass before any order is placed.

Then record the decision together with the assumptions that produced it. The next buyer will need to know whether the winning bid included freight and bulbs, and that is not something anyone will remember in eighteen months.

If you want the landed-cost mechanics rather than the scoring method, How to Build a Landed Cost Model for a Lighting Order goes through the calculation.

What a scoring sheet cannot tell you

A matrix compares what is written down, which means it rewards suppliers who document well as much as suppliers who perform well. That bias is real and worth stating explicitly. A small factory with excellent output can score badly simply because nobody there writes things down.

The correction is to test the top two rather than to refine the sheet. Place a small mixed-SKU order with each and compare what arrives, how it is packed, whether the paperwork matches the carton, and how quickly a question gets answered. Those four observations will separate two close bids faster than another scoring pass.

Record what the test showed next to the scores. It becomes the baseline for the next evaluation and stops the team from relitigating the same decision every season.

Finally, recognise that the sheet reflects the specification you wrote. If the top two bids sit within a point of each other, the likelier explanation is a specification too loose to discriminate, not two equally excellent suppliers.

Frequently asked questions

How many criteria should a scoring sheet have?

Enough to cover cost, compliance and service, and no more. Beyond roughly eight criteria the scores stop discriminating and the sheet becomes administration.

Should unit price carry the heaviest weight?

Usually, but rarely more than half. Programmes fail on serviceability, documentation and replenishment far more often than they fail on the initial price.

What if two suppliers score within a point of each other?

Test them. Place a small mixed-SKU order with each and compare what actually arrives, how it is packed, and how the paperwork reads.