Restaurant Vendor Scorecard: How to Evaluate Your Food Suppliers Every Quarter

A restaurant vendor scorecard is a one page summary, per vendor, of how often they delivered what you ordered, when they said they would, at the price you expected. You build it from records you already have: your purchase orders, the invoices you marked at the back door, and your log of shorts and credits. Once a quarter, you fill it in, look for the one or two numbers that are off, and decide what to do about each vendor: leave it alone, talk to the rep, move a few items, or line up a backup.

The point isn't to grade vendors for the sake of it. It's to replace "I feel like the produce guy is always late" with "5 of the last 13 Saturday deliveries came after the window." The second sentence gets a very different response from a rep.

Scoring vendors sits near the end of the purchasing loop, after orders are placed, received, and closed out. If you want to see how it connects to counts, pars, and ordering, our complete guide to restaurant purchasing and inventory management walks through every step.

In this guide

• What to measure on a vendor scorecard
• Where the numbers come from
• A sample restaurant vendor scorecard
• Worked example: three vendors, one quarter
• How to run a quarterly vendor review
• What to do with the results
• What chefs and line cooks can add
• Common mistakes
• FAQ

What to measure on a vendor scorecard

Keep it to a handful of measures you can actually pull from your records. Each one answers a question you'd otherwise argue about from memory.

A few notes on the definitions, because small differences change the numbers:

• Count lines, not cases, for fill rate. One missing case of your top protein matters more than one missing case of napkins, but counting lines keeps the math simple and makes every item visible. If you'd rather count cases, use the same method every quarter.
• Count substitutions separately from shorts. A sub you accepted still isn't what you ordered, and a vendor that quietly subs a different brand every week has a different problem than one that shorts.
• "On time" means the window you agreed on. If your vendor promised 7 to 10 AM, a 10:40 AM delivery is late, even if it's still before lunch. Write each vendor's window down before you start scoring against it. A truck that shows up before anyone is there to receive it is worth noting too.
• A billing error is anything that needed a credit or a correction: billed for a case that was shorted at the door, a price that doesn't match the quote, or the wrong pack size.

Where the numbers come from

You don't need special software for this. You need three records, kept consistently for the whole quarter:

1. Your purchase orders. What you asked for, from whom, on which day, and at what expected price. If you keep a PO log with a line for each item, you already have the "ordered" side of every measure. Our guide on how to track restaurant purchase orders shows how to set one up on paper or in a spreadsheet.
2. Your marked invoices. What actually arrived, what was short or subbed, what was refused, and what time the truck came. The only change most kitchens need to make is to write the arrival time on the invoice when they sign it. Without that, you can't score on time delivery.
3. Your short and credit log. Every credit you requested, when you asked, and when the credit memo arrived. That's where rep follow through comes from.

If those three records are patchy, the scorecard will be too. That's fine for the first quarter. Score what you have, note what's missing, and fix the gaps going forward. A scorecard built from half a quarter of good records is still more useful than one built from memory.

A sample restaurant vendor scorecard

Here's a blank version. One column per vendor, one row per measure, plus the threshold you've set for each. Copy it into a spreadsheet or onto a sheet of paper.

The last two rows matter most. "Biggest single problem" forces you to look past the averages to a specific item, day, or habit. "Action this quarter" makes sure the review ends in a decision, even if the decision is "nothing, they're fine."

I'd skip combining everything into one weighted score. A vendor with a 92 out of 100 tells you nothing about whether the problem is late trucks or bad invoices, and those need completely different fixes.

Worked example: three vendors, one quarter

This is an illustrative example. The restaurant, vendors, items, prices, and results are made up to show the method. They aren't typical figures for any distributor.

A restaurant uses three main vendors. It keeps a PO log, writes delivery times on invoices, and logs every credit request. At the end of the quarter (13 weeks), the GM pulls everything together.

The setup:

• Broadline distributor: delivers twice a week, about 40 lines per order, 7 to 10 AM window
• Produce vendor: delivers Tuesday, Thursday, and Saturday, about 14 lines per order, 6 to 9 AM window
• Meat purveyor: delivers twice a week, about 8 lines per order, 8 to 11 AM window
• The restaurant's own thresholds: 97% filled as ordered, 95% on time, 95% invoice accuracy, and 80% of issues resolved within two business days

Step 1: count from the records

Step 2: calculate the rates

Filled as ordered = (lines ordered − shorted − subbed) ÷ lines ordered

Broadline:  (1,040 − 31 − 18) ÷ 1,040 = 991 ÷ 1,040 = 95.3%
Produce:    (546 − 11 − 4) ÷ 546     = 531 ÷ 546   = 97.3%
Meat:       (208 − 9 − 0) ÷ 208      = 199 ÷ 208   = 95.7%

On time = (deliveries − late) ÷ deliveries

Broadline:  25 ÷ 26 = 96.2%
Produce:    33 ÷ 39 = 84.6%
Meat:       26 ÷ 26 = 100%

Invoice accuracy = (invoices − invoices with an error) ÷ invoices
(one invoice per delivery)

Broadline:  23 ÷ 26 = 88.5%
Produce:    38 ÷ 39 = 97.4%
Meat:       24 ÷ 26 = 92.3%

Step 3: price change on a fixed basket

To see whether a vendor is getting more expensive, price the same items at the same quantities in both quarters. Holding quantities fixed means a busy quarter doesn't look like a price increase. Here's the broadline's basket, using this quarter's quantities for both:

Change:  $9,499.50 − $9,161.70 = $337.80 more for the same five items
Percent: $337.80 ÷ $9,161.70 = 3.7%

The GM runs the same check on the produce and meat baskets: produce is up 6.2% and meat is up 1.1%. Produce prices move with the season, so a quarter over quarter jump there means less than the same jump on dry goods or paper. Comparing against the same quarter last year is fairer for produce, if you have it.

Step 4: the filled in scorecard

Bold marks anything below the restaurant's threshold. At first glance, every vendor has a problem. That's the point where most people stop and draw the wrong conclusion. The next step is to look underneath each bold number.

Step 5: find the biggest single problem

Broadline. The fill rate is below target, but the short and sub lines are spread across lots of different items, with no one item standing out. The bigger issue is billing: 3 invoices needed corrections. Two billed for cases that were marked short at the door, and one charged an old, higher price on fryer oil. Combined with 3 credits that took longer than two business days, that's a vendor whose back office needs a conversation. The basket increase is led by fryer oil (+7.1%) and to go containers (+5.7%).

Produce. Quality and fill are fine. Five of the six late deliveries were on Saturdays, out of 13 Saturday deliveries. The Tuesday and Thursday trucks were late once in 26. This isn't a vendor problem so much as a route problem.

Saturdays:          8 on time out of 13 = 61.5% on time
Tuesday + Thursday: 25 on time out of 26 = 96.2% on time

Meat. Perfect on time record, no subs, nothing refused. But 6 of the 9 short lines were the same item: hanger steak, ordered on 13 Thursday orders and shorted on 6 of them. The two slow credits were both hanger steak too. Take hanger out and the picture changes:

Lines without hanger steak:     208 − 13 = 195
Shorts without hanger steak:    9 − 6 = 3
Filled as ordered without it:   (195 − 3) ÷ 195 = 98.5%

One item is dragging down an otherwise excellent vendor.

Step 6: one action per vendor

Nobody gets fired. Each vendor gets one specific conversation, and the GM knows exactly what to check next quarter.

How to run a quarterly vendor review

For a restaurant with three to five main vendors, this usually takes an afternoon once the records are in decent shape.

1. Pull the quarter's records for each vendor: POs, marked invoices, and the short and credit log.
2. Count and calculate the measures on the scorecard. A spreadsheet with one tab per vendor makes the second quarter much faster than the first.
3. Compare against your thresholds and mark anything below.
4. Dig into each marked number. Sort by item, by delivery day, and by who received the order. Look for the one item, day, or habit behind the number.
5. Ask the kitchen about quality and anything the records don't show (see what chefs and line cooks can add below).
6. Pick one action per vendor and write it on the scorecard.
7. Talk to each rep about anything you marked. Bring the numbers.
8. Keep last quarter's scorecard next to this one. A trend over two or three quarters tells you much more than any single quarter.

Watch your sample sizes

A quarter isn't a lot of deliveries. With 26 deliveries, one late truck moves on time delivery by almost 4 points (1 ÷ 26 = 3.8%). Before you act on a change, look at the actual count. "Went from 100% to 96%" sounds worse than "one late truck."

Quarterly is a good default, not a rule

Quarterly works for established vendors. For a new vendor, review monthly for the first few months, since that's when problems with specs, routes, and billing usually show up. For a vendor you're already worried about, a monthly check on the one measure you care about is fine too.

What to do with the results

Most results call for a conversation, not a change of vendor. Here's a rough ladder, from least to most disruptive.

1. Do nothing

If a vendor clears your thresholds and nothing stands out underneath, tell them. Reps hear about problems all the time. Hearing "your Saturday truck has been great all quarter" builds the kind of relationship that helps when you need a favor at 4 PM on a Friday.

2. Talk to the rep

Bring specifics: dates, PO numbers, items, and counts. "Six of thirteen Thursday orders came without hanger steak" gives the rep something to take to their buyer or warehouse. "You guys are always shorting us" doesn't. Ask what's causing it and what they can change. Then check next quarter whether it changed.

If nothing changes after a conversation or two, it's reasonable to ask whether there's a sales manager or account manager above your rep who should see the numbers.

3. Fix it on your side

Some problems are partly yours. If a vendor's late deliveries are all on one day, maybe the order or the delivery day should move. If shorts happen mostly on orders placed close to the cutoff, ordering earlier might help. If the same items keep getting subbed, your spec may be something the vendor struggles to stock, and it's worth asking what they carry reliably.

4. Shift a few items

When one item is the problem, move that item, not the whole vendor. In the example, the meat purveyor is excellent except for hanger steak. Moving hanger to another source keeps everything that works.

Before you move items, check the effect on both orders. Pulling items off one vendor makes that order smaller, and adding them to another vendor might mean another delivery to receive. Our guide on vendor order minimums and delivery fees covers what happens when a lighter order drops under a minimum.

5. Add a backup vendor

For critical items, the ones where running out means 86ing a dish, a backup is worth setting up whether or not a vendor is struggling. A scorecard that shows a pattern of shorts on one item is the signal to stop putting it off. Our guide to handling shorted items and substitutions covers how to set up backup sources and cover a gap when a short happens.

6. Get quotes and consider switching

If a vendor misses on the same measure for two or three quarters after you've raised it, it's time to price the alternative. Compare the same items at the same spec, including fees and yield, as covered in comparing restaurant vendor prices. Then weigh what you'd gain against what changes: another account to manage, new delivery days, and the risk that the new vendor has problems of its own. If you're thinking about consolidating categories onto one distributor or splitting them off, see broadline vs specialty distributors.

Make one switch at a time, then score the new vendor monthly until you trust it.

What chefs and line cooks can add

The records cover quantities, times, and prices. They don't capture much about quality, and quality is often why you picked a vendor in the first place. The people cooking with the product see things the invoice never will:

• Product that doesn't hold. Herbs from one vendor that wilt a day sooner, or fish that doesn't make it to the third day.
• Inconsistent specs. Chicken breasts that vary wildly in size from case to case, or avocados that are never ripe on the day you need them.
• Subs nobody flagged. A different brand of fryer oil that foams sooner, or a different cheese that melts differently.
• Yield changes. If a case of something suddenly yields less after trim, that's a hidden price increase.

Ask for specifics with dates, and keep a simple quality note sheet near the walk in for the quarter. "Case of cilantro on 9/12 was slimy by day two" is something you can take to a rep. "Their herbs are bad" isn't.

Common mistakes

• Scoring from memory. The vendor that annoyed you last week will always feel like the worst one. Use the records.
• Acting on averages without digging in. A low fill rate driven by one item calls for a very different fix than one spread across fifty items.
• Comparing vendors that do different jobs. A produce vendor with three deliveries a week and a dry goods vendor with one aren't being asked to do the same thing. Compare each vendor against your thresholds and its own past quarters.
• Not writing down the delivery window. If you never agreed on a window, you can't call a delivery late.
• Forgetting to count credits that never came. A short that was marked but never credited is money you paid for product you didn't get.
• Overreacting to small numbers. One late truck out of 26 is not a pattern.
• Keeping the scorecard to yourself. Share the relevant numbers with each rep. A vendor can't fix what they don't know you're measuring.
• Switching vendors over one bad quarter. Raise it, give them a quarter to fix it, and then decide.

FAQ

What should a restaurant vendor scorecard include?

At a minimum: how often orders were filled as ordered, the short and substitution rates, on time delivery against the agreed window, invoice accuracy, price changes on your top items, items refused for quality, and how quickly the rep resolved problems. Add a row for the biggest single problem and the action you'll take.

How do you evaluate a food supplier?

Use your own records instead of impressions. Compare what you ordered (your purchase orders) with what arrived and when (your marked invoices) and what happened afterward (your credit log). Calculate a few simple rates per vendor, look for the specific item or delivery day behind any weak number, add the kitchen's notes on quality, and talk to the rep about anything that falls short.

What's a good fill rate for a restaurant distributor?

There isn't a reliable industry standard for independent restaurants. Set a threshold based on what your kitchen can absorb, and track the trend. A missing case of a critical protein matters more than a percentage point on dry goods, so always check which items are behind the number.

How often should I review my vendors?

Quarterly works well for established vendors. Review new vendors monthly for their first few months, and check any vendor you're worried about monthly on the one measure that concerns you.

Should I tell my rep I'm tracking their performance?

Yes. Share the numbers that matter for their account, with dates and PO numbers. Most reps would rather fix a specific problem than lose an account over a vague one, and a rep who knows you keep records tends to follow up on credits faster.

Do I need software to keep a vendor scorecard?

No. A spreadsheet with your PO log, a short and credit log, and invoices with delivery times written on them is enough. The hard part is keeping the records consistently for a full quarter, not the math.

Keeping the ordering side of the scorecard in one place

A scorecard is only as good as the order records behind it. Proteus Kitchen keeps every purchase order you build, with its vendor, dates, items, quantities, and the estimated price for each line, in one list you can filter by vendor and date range. Each vendor has its delivery days, order deadlines, contact information, and a notes field, which is a good place to record the delivery window you agreed on and how that vendor handles credits. The weekly report shows the number of orders and estimated spend by vendor.

Proteus doesn't score vendors for you, and it doesn't record what was received against each order, so shorts, subs, delivery times, and credits still get marked on the invoice and in your short log. It doesn't place orders with vendors either. It generates each order as text to send through the vendor's portal, a text, or an email, and it doesn't need a POS integration.

If you want somewhere to write a backup vendor next to each critical item, the free restaurant order guide template has room for it. If your vendor list or order guide is already in a spreadsheet, Free Restaurant Setup covers bringing it into Proteus, and there's a 30 day free trial below.

Next step: score one vendor this week

1. Pick your biggest vendor and pull the last 13 weeks of orders and invoices.
2. Count deliveries, lines ordered, shorts, subs, and invoices that needed a correction.
3. Calculate filled as ordered and invoice accuracy. Look for the one item behind any weak number.
4. Start writing the arrival time on every invoice, so next quarter you can score on time delivery too.
5. Bring one specific finding to your rep.