Perfect Order Rate (the only quality metric SMB e-commerce needs).

What POR actually measures

Perfect Order Rate (POR) is one of the foundational metrics in the SCOR Digital Standard from ASCM, formally SCOR RL.1.1. It measures the percentage of orders that are perfect: delivered on time, in full, undamaged, and with accurate documentation.

Most e-commerce stores track on-time, in-full, damage, and returns as four separate numbers and never roll them up. That made sense in 1995. It stopped making sense the day Walmart started asking suppliers for a single number, which they called Perfect Order Rate. The customer doesn’t experience your operation as four metrics. They experience one shipment, and the shipment is either perfect or it isn’t.

If 95% of your orders ship on time, 95% are in full, 95% arrive undamaged, and 95% have correct documentation, that’s not 95% perfect orders. That’s 81%. POR is multiplicative, not additive. That one fact is what separates people who use the metric well from people who quote it without understanding it.

The formula

The SCOR-standard formula is the product of four independent factors:

POR = on_time_pct × in_full_pct × damage_free_pct × doc_accuracy_pct

Each factor is the percentage of orders meeting that criterion, expressed as a decimal between 0 and 1. The product is your POR, also as a decimal between 0 and 1.

The four factors, defined

• On-time, delivered by the date promised at checkout. Measured against the carrier’s delivery confirmation timestamp, not the ship date.
• In-full. Every line item the customer ordered is in the package, in the right quantity. Partial fulfillments and substitutions count against you.
• Damage-free, arrives in saleable condition. Inferred from “damaged” / “wrong item” return reasons and customer complaints; subtract from total to get the rate.
• Documentation accuracy, invoice, packing slip, and shipping label match the order exactly. Wrong recipient, wrong product description, wrong VAT line all count against you. Less visible in DTC than in B2B but real.

Why we multiply (not average)

If a customer’s order arrives on time but with the wrong item, did they have a perfect experience? No, even though three of four factors were perfect. The order failed.

An order is only perfect if every factor is met. That’s a logical AND. In probability terms, when independent events must all be true, you multiply their probabilities. That’s why POR is a product, not an average.

Most stores that try to compute their own POR average the four factors instead of multiplying. That gives a number that looks generous and isn’t useful. Don’t do that.

How to measure each factor on Shopify

1. On-time delivery

Shopify exposes fulfillment.delivered_at and the original order timestamp. The promised delivery date is harder, it lives in your shipping configuration or in the carrier’s tracking estimate.

Practical approach: take your stated delivery promise (e.g., “5 business days for ground shipping”), add it to the order date, and compare against the actual delivered timestamp. If your store doesn’t make explicit promises, use the carrier’s estimated delivery date as the reference.

Use carrier-confirmed delivery, not “out for delivery” or “shipped”. Customers measure delivery by when the box reaches their door.

2. In-full

An order is in-full when every line item shipped, in the quantity ordered. In Shopify terms: every line_item.fulfilled_quantity equals line_item.quantity on the order.

Be careful with split shipments. They’re still in-full if all line items eventually ship in correct quantities. Set a “in-full deadline” (e.g., within 14 days of order) so backorders don’t count as in-full forever.

3. Damage-free

You can’t directly measure this from Shopify data. The proxy: orders with returns or refunds whose reason codes indicate damage, wrong item, or quality issue. Loop Returns, ReturnLogic, and other returns platforms expose return-reason codes; Shopify’s basic return system does not consistently.

Practical formula:

damage_free_pct = 1 - (damage_returns + wrong_item_returns) / total_orders

If you don’t have reason codes, use a conservative proxy: half of all returns. That tends to overstate damage rate, which makes your POR conservative, fine direction.

4. Documentation accuracy

This is the hardest to measure programmatically. In B2B and wholesale operations it shows up as invoice corrections; in DTC it tends to be invisible until a customer escalates.

Pragmatic approach: assume 99% accuracy as your baseline unless you have evidence of systematic invoice errors. Treat improvements here as a low priority unless your customer-support tickets are full of “wrong on the receipt” complaints.

What’s a good Perfect Order Rate?

SCOR-standard benchmarks vary by industry. For e-commerce specifically:

A 75% POR means one in four customers got a shipment that disappointed them on at least one dimension. That number is a leading indicator. Refunds and chargebacks land 30 to 60 days later. Churn shows up in the next quarter. Bad reviews live forever on Trustpilot.

How to improve POR

POR is multiplicative, so the highest-leverage move is always to find your weakest factor and improve it. Pulling a 92% factor up to 95% does more for POR than pulling a 99% factor up to 99.5%.

If on-time is your weak link

• Audit carrier scorecards. One slow carrier on one lane often drags the whole metric down.
• Look at the Packing → Shipped transition specifically. Most “late” orders are late because they sat in packing too long, not because the carrier was slow.
• Check for warehouse dead zones, hours when no shipments go out. Friday afternoons and Monday mornings are common.

If in-full is your weak link

• Audit your inventory accuracy. Stockouts that aren’t reflected in Shopify cause overselling.
• Look at split-shipment rates. Multi-warehouse strategies that aren’t tuned create false in-full failures.
• Check your supplier scorecards if you drop-ship. Slow suppliers are usually invisible in your in-full number until you decompose it.

If damage-free is your weak link

• Audit packaging. Specific products with specific carriers often have specific damage rates.
• Audit return reasons by SKU. Damage rates aren’t uniform. They cluster on a small subset of products.

Common mistakes

Averaging the four factors instead of multiplying

Already covered, worth saying twice. If you average, you’re not computing POR. You’re computing a feel-good number.

Using ship-confirm SLA instead of delivery SLA

Many 3PLs report against ship-confirm SLA, did they hand the package to the carrier on time. That’s not on-time delivery. A 3PL can be 100% aligned with on ship-confirm while 25% of customers receive late deliveries because the carrier missed.

Computing POR weekly when your sample size is too small

POR’s confidence interval depends on volume. Below ~200 orders/week the metric becomes noisy. Aggregate to monthly or rolling 30-day windows for stores doing under 1,000 orders/week.

Treating POR as a static target

Your POR floor is 100%. Every order should be perfect. POR is a tracking metric, not a target. The question is “what’s your POR trend, and what’s the dollar impact of the gap to 100%?”, not “did we hit 87% this week?”

The takeaway

Perfect Order Rate is one number that tells you, every day, whether your operation is keeping its promises to customers. It is multiplicative, so it does not flatter you. It has been a standard inside supply-chain orgs since 1995, and the math works the same on a Shopify store as it does on a Walmart pallet.

Compute it by hand once and the conversation with your ops team changes. Compute it inside Instirio and every drift, regression, and contributing detector lights up the moment a factor starts slipping. The math is right by default. The work for you becomes deciding what to fix first.

Sources: ASCM SCOR Digital Standard; SCOR RL.1.1 metric definition; Instirio detection methodology.