Right before Black Friday 2023, a seller I know moved his inventory from a small self-run warehouse in California to a 3PL in Arizona. The new place charged 40 cents less per pick-and-pack, and he figured he'd save tens of thousands a year. Then, in the third week of November, his hero SKU went out of stock on Amazon. Not because he was out of inventory — his pallets sat in the 3PL's receiving area for nine straight days, never checked in, showing as "in transit" in the system the whole time. He lost roughly a third of his expected holiday sales and collected a batch of one-star reviews. The money he saved on pick fees didn't even cover the off-Amazon ad spend needed to repair the listing.
I tell everyone this story, and I always end it the same way: the 3PL with the lowest unit price usually gives you the most expensive total bill. These five metrics come from twenty years of touring warehouses and switching 3PLs. Every one of them converts directly into money. Ask all five before you talk price with your next provider.
Arrived at the Warehouse Doesn't Mean Sellable: Receiving-to-Stock Time
This is where first-time 3PL users get burned. The contract says "receiving," and you assume the job is done when the truck unloads. In reality there's a wall between the dock door and a sellable SKU: unloading, counting, QC, system check-in. Some warehouses run receiving with two or three people, and in peak season the staging area turns into a mountain. Freight sitting there for a week is not unusual.
Don't ask "are you fast at receiving." Ask three specific questions: how many hours from trailer arrival to unload-and-count completion; how many hours from count completion to the SKU showing sellable in the system; and how much those numbers stretch during peak season — with a cap. Then don't just take the sales rep's word for it. Ask to see the actual receiving-to-stock numbers from the last three months. A serious 3PL has that report. One that can't produce it isn't managing the metric.
How do you write this into the contract? Put "system check-in completed within X hours of receiving" into the SLA, with a per-day credit against storage or service fees when they miss it. Never accept words like "prompt" or "timely" — they're unenforceable when there's a dispute. And do the cheapest due diligence there is: before the full cutover, send a small test shipment through the entire flow and time it from dock arrival to sellable. It beats a hundred promises.
One Mispick Costs More Than Freight: Pick Accuracy and On-Time Ship Rate
These two are the lifeblood of a 3PL, and they're the easiest to hide behind a low unit price. One mispick costs you return shipping, reshipment freight, customer service time, and possibly a one-star review. Reviews don't go away, and they tax your conversion rate — a hidden levy on every future order.
Pick accuracy first. Good warehouses run at 99.5% or better; bad ones can't hold 98%. That 1.5-point gap sounds small until you do the math: at 10,000 orders a month, it's 150 wrong packages. Verification is straightforward — ask for the mispick and complaint rates for the last 90 days, exported from the system, not quoted by a salesperson. Pay attention to the definition they're using. If you can, visit and watch the verification step: is there a second scan? Weight checks? A warehouse that verifies by eyeball alone will see accuracy collapse the moment peak-season volume hits.
Then the on-time ship rate, where the real question is the order cutoff time. "Same-day shipping" means two different businesses when one warehouse cuts off at 5 PM and another at noon. Get the cutoff time in writing, how many hours after cutoff orders must leave the dock, and whether there are multiple outbound waves per day. Put the cutoff and the "orders placed before cutoff shipped same day" rate into the contract with a remedy clause. You may never collect the credit — that's not the point. The point is making them treat the number as real.
System Integration Is the Dividing Line Now
Ten years ago, system integration was a nice-to-have when choosing a 3PL. Today it's the line between surviving and not. Your orders flow in from Shopify, Amazon, TikTok Shop; inventory levels, tracking numbers, and return statuses have to sync back in near real time. Anything that depends on someone exporting a spreadsheet will break, and it will break slowly enough that you notice too late.
Ask technical questions in the meeting: which integration methods do you support — real-time API or a once-a-day FTP dump? How often does inventory sync, in minutes? Is there a sandbox for testing before go-live? How long does a typical integration take, and who does the work on your side? If the answer is "we have an Excel template, just send it daily," walk away. That warehouse cannot handle a multi-channel business, and the integration is the first thing that falls over on a spike day.
There's a trap that only surfaces when you leave: data portability. When the relationship sours and you need to move, can you export your full order history and lot-level inventory data? Some 3PLs run on old systems that can't even produce a complete outbound detail file, which means switching providers equals a data blackout. Get the export format and turnaround time into the contract before you sign, so the breakup doesn't turn ugly.
If You Can't Read the Bill, Don't Sign: Billing Transparency
I've read dozens of 3PL rate cards, and I'd say nine out of ten have a catch — not fraud, just a fee structure so tangled you can't compute your own total. Pick-and-pack and storage look cheap, and then the accessorials run twenty lines deep: receiving fees, putaway fees, labeling, cycle counts, return processing, system access, packaging materials, oversize surcharges. Forty cents saved on the pick rate, a dollar twenty added back in fees — I've seen that movie more than once.
The fix is one sentence: make them build a sample invoice from your real volume. Hand over last month's order profile — order count, SKU count, average units per order, return volume — and have them price every line item. Compare providers on that same-profile total. That is your real unit price. Also ask for a redacted sample invoice from a real client and look at the accessorial share: if fees beyond the headline rates exceed roughly a third of the total, the rate card is bait.
Lock two things into the contract. First, a rate-change notice period — at least 30 days in writing, no surprise increases. Second, new fee line items require your written approval. Most billing disputes I've seen trace back to "we've always charged for that, we just never mentioned it."
No 3PL is perfect; it's all trade-offs. Fast check-in isn't cheap, and the warehouse with the best systems might have the earliest cutoff. Rank by your stage: if you're doing a few thousand orders a month and growing, put integration and receiving-to-stock time first. At tens of thousands of orders a month, pick accuracy and billing transparency start to matter more.
One thing you can do tomorrow: score your current 3PL against these five metrics. Whichever one you can't answer or can't get data for is your next peak season's landmine. And if you're planning a switch — don't do it in December. I've watched too many sellers move warehouses mid-holiday, with freight floating between two facilities for two weeks, belonging to neither.
