Vendor Scorecards That Move Supplier Volume

Most brands know which supplier is annoying. Very few can say which supplier is expensive. Those are different questions, and only the second one is answerable with data.

A vendor scorecard is not a governance ritual. It is a mechanism for moving volume toward the suppliers who cost you least in total, including all the costs that never appear on their invoice.

The five metrics that carry the weight

Resist the urge to measure fifteen things. Five is enough, and each one has to map to a rupee cost you actually incur.

• On-time delivery. Percentage of purchase order lines received inside the committed window. Use a window, not a date: on or before the promise date, and not more than three days early, because early receipts eat storage and cash. Measure against the original promised date, never the revised one.
• Fill rate. Quantity received divided by quantity ordered, at line level, on the first delivery. A supplier who ships 70 percent on time and completes the rest a week later is a stockout waiting to happen. Track the combined OTIF figure, on time and in full, because that is what your planning actually depends on.
• Defect rate. Rejections at inbound inspection plus field failures traced back to that supplier, expressed per hundred or per thousand units depending on your volumes. Attach the rework cost and the attributable customer returns so the number has weight in the room.
• Responsiveness. Median hours to acknowledge a purchase order, and median hours to respond to a quality escalation. This sounds soft. It is not. Slow acknowledgement is the leading indicator of a missed delivery, usually by ten to fourteen days.
• Price stability. Number of unplanned price revisions in the period, plus the variance between quoted price and final invoiced price including freight and incidentals. A supplier who is 3 percent cheaper on quote and 4 percent dearer on invoice is not cheaper.

If you buy anything regulated or perishable, add a sixth. Documentation completeness. Missing certificates of analysis, batch records or e-way bills cost you more in blocked stock than most defects do.

Collecting the data without buying a system

You do not need a vendor management module. You need three columns added to records you already keep.

• In the purchase order register, add promised date, actual receipt date and quantity received. Your Tally, Zoho or Busy data already carries two of the three.
• At goods receipt note stage, add a rejected quantity field and a one word reason code. Six codes cover almost everything: damage, short, wrong item, spec fail, documentation, other.
• Keep one shared sheet where whoever owns the supplier logs escalations with a timestamp for raised and for closed.

That is the whole data layer. One person spends about 90 minutes a month pulling it into a pivot. For a brand with 12 to 30 active suppliers this is genuinely all it takes. Buy software when you cross 60 suppliers, not before.

Two collection rules protect the numbers. Record the promise at the moment the purchase order is accepted so revisions cannot be backfilled. And do not let the person who owns the supplier relationship be the same person who enters the defect data.

Weighting, banding and the score

Weight by what hurts you. A sensible default for a consumer brand is on-time delivery 25 percent, fill rate 25 percent, defect rate 25 percent, price stability 15 percent and responsiveness 10 percent.

Then shift the weights to match your actual failure mode. If you sell on quick commerce and a stockout costs you shelf position, push the two availability metrics to 60 percent combined. If you are in personal care where a bad batch means a recall, defect rate carries more.

Convert each raw metric to a 0 to 100 score against a stated target, not against your best supplier. Ranking suppliers against each other simply rewards your least bad option. Targets that hold up in Indian consumer supply chains:

• On-time delivery 95 percent
• Fill rate 98 percent
• Inbound rejection under 1 percent
• Purchase order acknowledgement within 24 hours
• Zero unplanned price revisions in the quarter

Band the total score. Eighty five and above is preferred. Seventy to 84 is approved. Below 70 is on watch with a documented improvement plan and a review date.

Running the quarterly business review

Quarterly, 60 minutes, same agenda every time. Send the scorecard 72 hours ahead so nobody spends the meeting reading it.

• Ten minutes on the score and the four quarter trend
• Fifteen minutes on the two worst line items only, with the supplier explaining root cause rather than you explaining the number
• Ten minutes on your forecast for the next two quarters, which is the part they came for
• Fifteen minutes on their asks, including anything you do that makes their life harder
• Ten minutes on actions with owners and dates

The fourth item matters more than brands expect. A large share of supplier misses trace back to buyer behaviour: artwork approved four days late, purchase orders released in a lump on the last day of the month, specification changed after material was already procured. A scorecard that never looks in the mirror gets quietly ignored.

Use it to move volume, not to complain

A scorecard with no consequence is a newsletter. Attach volume to it before the first review and say so up front.

• Publish the allocation rule. On any part with two qualified suppliers, the preferred band takes 70 percent of next quarter volume and the approved band takes 30 percent. Band changes, allocation changes, automatically.
• Give the watch band one quarter and a written plan with three specific actions. If the score does not move, shift 20 percent of their volume and tell them exactly why.
• Reward the top band with something they value more than praise. Faster payment, a longer commitment, or first refusal on a new SKU.
• Never move an entire part to one supplier on the back of one good quarter. Dual sourcing is the reason the scorecard has teeth at all.

Six months in, the useful output is not the ranking. It is that suppliers start managing to the metric before the review happens, because they can predict what the meeting will say. That is the whole point of building it.