Restaurant Supplier Evaluation: A Scorecard Framework for Multi-Site Groups Deciding Which Suppliers to Keep

What a Supplier Scorecard Measures That Gut Feel Cannot

A restaurant supplier evaluation scorecard rates every supplier on the same fixed criteria, then turns those scores into one decision: keep as primary, keep as backup, or drop. Instead of a manager's memory of who annoyed the kitchen last month, each supplier carries one comparable number that every location can see and every buyer can defend.

At a single site, gut feel is survivable. The head chef knows which produce supplier short-delivers on a Friday and quietly orders extra. Across a group of 18 active suppliers feeding 12 locations and roughly $2.4M of annual purchasing, that knowledge does not scale. One branch keeps a supplier the group should have dropped; another drops a supplier that performs well everywhere else. Nobody is comparing the same things, so nobody is really deciding, they are just reacting.

A workable scorecard fixes six criteria and weights them by what actually costs you money. A sample weighting: fill rate and order accuracy at 25%, delivery reliability at 20%, price stability at 20%, quality and returns at 15%, lead time at 10%, and communication and documentation at 10%. The exact weights are yours to set, but two rules hold: price is never the only axis, and the criteria stay identical for every supplier so the scores mean the same thing.

The uncomfortable part is that some suppliers cannot be scored on paper at all. It is common for half of purchasing volume to run through informal vendors with no invoices and messaging-app orders, and a scorecard has to acknowledge them rather than pretend they do not exist. The point of the framework is not a perfect number, it is a consistent one, so that a decision to keep or drop a supplier survives a change of manager, a new location, and a difficult quarter. For the wider system this sits inside, see how multi-site operators build a restaurant supplier management process end to end.

Where a Supplier Looks Fine on Price but Fails on Fill Rate

Fill rate is the share of what you ordered that actually arrived, in full, on the delivery it was promised on. It is the criterion price sheets never show, and it is usually the one that quietly costs the most. One multi-site group's back-of-house had to raise a credit note for an item a supplier never delivered at all, and the shortfall only surfaced days later when someone reconciled the paperwork, not at the door when the menu item was already running low.

Scored across a portfolio, the differences are stark. A strong primary supplier lands a 98% fill rate on a 2-day lead time. A middling one sits at 91%. A weak one runs 82% with a 5-day lead time and needs a credit note on 6% of its deliveries. On the quoted price sheet all three can look interchangeable, but the 82% supplier is forcing last-minute substitutions, emergency orders from a backup, and hours of staff time chasing credits the group is owed.

That is why fill rate and delivery reliability carry the two heaviest weights in the sample scorecard. To score them without a clipboard, you read them straight from goods-received data: what was ordered on the purchase order versus what was signed for on delivery, and how often the two disagree. A supplier's returns and credit-note history is the same signal from the other side, which is why tracking receiving discrepancies consistently is what makes the fill-rate score trustworthy rather than anecdotal.

When a Cheaper Headline Price Hides Silent Cost Creep

Price belongs on the scorecard, but the number that matters is price stability, not the headline rate on the day you signed. A multi-site group discovered that unnoticed supplier price increases had been slipping through paper-based goods-received processes for weeks, with no automated purchase-order-to-invoice check to catch the drift. The supplier still looked cheap on the original quote while quietly rewriting it one invoice at a time.

The math is unforgiving at volume. A supplier carrying $600,000 of annual purchasing with a 4% price drift that nobody flags costs the group $24,000 a year, enough to outweigh whatever headline discount won them the business. A scorecard that only records the quoted price rewards exactly the wrong supplier: the one who bids low and claws it back on the invoice.

The score input for price stability is invoice-to-purchase-order variance: how often, and by how much, the price billed departs from the price agreed. Received against the order line by line, a creeping supplier shows up as a pattern of small overages rather than one obvious spike, which is what let the drift run for weeks in the first place. For the deeper method here, see how groups benchmark supplier prices across sites so a rise at one branch is visible against the same item everywhere else.

Primary, Backup, or Drop: Turning Scores Into a Decision

The scorecard only earns its keep when the total score maps to an action. A simple rule works: score 80 or above out of 100 and the supplier stays primary; 65 to 79 and they are kept as a backup rather than a first call; below 65 and you either renegotiate against the specific criteria dragging them down or drop them. In the running sample, Greenfield Produce scores 92 and stays primary, Harbour Provisions scores 78 and becomes a documented backup, and Summit Meats scores 61, which triggers a renegotiation on fill rate before anything else.

That distinction is not cosmetic. A finance lead at one group found that leaving backup suppliers unranked in the product master inflated recipe costs, because average-costing logic blended a rarely-used backup's price into every plate. Designating a clear primary versus backup per ingredient is what keeps costing accurate and what makes the scorecard's output real rather than a filed document. Informal vendors get scored on what you can see, price and remembered reliability, kept as backups where they genuinely earn it, and steadily pushed toward documented ordering so they can be scored properly next quarter.

None of this needs a separate tracking spreadsheet, because the scores come from data the ordering process already captures. Supy holds per-branch delivery schedules and cut-offs and aligns purchase orders to them, so late and off-schedule deliveries are measurable rather than remembered. Invoice receiving flags price and quantity conflicts against the purchase order, which feeds the price-stability and fill-rate scores directly, and credit notes and returns are tracked with an audit trail, up to 5 approval levels sit over spend, and 75+ integrations pull the POS and catalog data underneath. You can see how the pieces fit on the Supy platform overview.

So the decision rule, stated plainly: keep a supplier as primary when they clear roughly 80 on the weighted score and hold fill rate in the mid-90s or better; keep them as a backup when they land in the 65 to 79 band, useful but not your first call; and renegotiate or drop them below 65, leading with the single criterion pulling the score down rather than a vague complaint. Re-score on a fixed cadence, quarterly is enough for most groups, so a supplier who has slipped is caught by the number before they are caught by a stockout on a Friday night.