Choosing a third-party logistics partner is a two-sided bet: you are wagering your delivery promise and your reviews on someone else's warehouse discipline, and they are wagering capacity on your forecast. Most brands evaluate providers by price per order and discover the real differences six weeks into peak season. This article gives you the dimensions that actually predict performance, the questions that expose them, and the red flags that should end the conversation.
Start from your order profile, not the sales pitch
Before contacting anyone, write down the profile a provider is actually agreeing to serve. The numbers that determine fit are unglamorous: SKU count and variant depth; average and peak orders per day; average parcel weight and dimensions; destination mix by country and region; channel mix across your own store, marketplaces and wholesale; and seasonality — how much your peak month exceeds your average. Two providers can both be excellent and be wrong for you: a facility built for pallet-in, pallet-out wholesale distribution will mangle single-parcel DTC work, and a boutique operation built for gift-quality unboxing may drown at your Black Friday volume.
Send the same written profile to every provider you shortlist. This single habit improves the quality of quotes more than any negotiation tactic, because it forces answers against your reality instead of their best case.
Nine dimensions that predict how the relationship will go
These are the dimensions worth probing, with the questions that separate real capability from brochure capability.
| Dimension | What to ask | What a red-flag answer looks like |
|---|---|---|
| Order profile fit | "Walk me through a day at my volume, peak included." | Vague reassurance; no questions asked back about your profile |
| Systems and integration | "Which warehouse system do you run, and what do I see in real time?" | Refuses to name the system or demo live stock levels |
| Picking accuracy | "How is a pick verified, and what happens when verification fails?" | "We're very careful" — no scan verification, no error process |
| Quality checks | "What is inspected inbound, and what is checked before pack?" | Nothing inbound; problems become your customers' discoveries |
| Pricing transparency | "Itemize inbound, storage, pick-pack, materials and shipping." | One blended number; "details at contract time" |
| Capacity headroom | "What is your peak plan for a client at 3x my volume?" | No written peak plan; staffing "handled as needed" |
| Exception handling | "Show me your exception queue: categories, owners, response times." | Exceptions handled by chat, no queue, no owners |
| Returns operations | "Walk me through a return from door to disposition." | Returns pile up ungraded; no recovery reporting |
| Communication model | "Who owns my account, and what is the escalation path?" | One WhatsApp contact, no ticket trail, no backup person |
Notice that only one row is about price. Price is discoverable from a quote; discipline is discoverable only from process questions, and it is discipline that determines whether the quoted price survives contact with your actual order flow. Our fulfillment operations page describes how we run each of these dimensions in practice, precisely so prospective clients have something specific to hold us against.
Red flags that should end the conversation
Some answers are not weaknesses to weigh — they are signals to walk away. End the evaluation if you encounter any of these:
- Refuses reference calls, or cannot produce a single client at a comparable volume and channel mix.
- Promises perfect accuracy. Every physical operation has an error rate; a professional measures it, publishes it internally and fixes root causes. A provider claiming 100% is either not measuring or not telling you the truth.
- No written service agreement covering turnaround, response times and error remediation — only verbal assurances.
- Storage or handling fees that appear only after you have moved inventory in, especially the months after the promotional onboarding rate expires.
- No disaster answer: ask what happened in their last major disruption — power, systems, carrier failure — and listen for whether a real post-mortem exists.
- Cannot explain how returns are inspected, graded and restocked, because returns will be a meaningful share of your operation whether you like it or not.
Comparing quotes on the same basis
Provider quotes are rarely comparable as written, so normalize them yourself. Build a simple monthly model at your real order profile: inbound per pallet or carton received, storage per unit or cubic unit occupied, pick and pack per order at your average item count, materials per parcel, and shipping rates sampled on your top lanes. Then stress the model in two directions — a slow month, where minimum storage or account fees may dominate, and a peak month at your realistic multiple, where surcharges and overtime live. The provider whose quote survives both months with transparent line items is almost always the cheaper one per year, whatever the headline rate said.
Ask every provider to price the same assumption set, and be suspicious of the outlier that is dramatically cheaper across every line. Fulfillment has real costs; a quote that hides them is deferred, not avoided. If you want a comparable baseline, request an itemized quote against your SKU and volume profile and use it as the template for the others.
The first thirty days set the tone
Onboarding is your first real data about the provider, so structure it deliberately. A sound sequence runs: SKU data mapped and every item barcoded; store integration connected and test orders run through the full loop; a first inbound received against a packing list with a written discrepancy report; a trial batch of real orders picked, checked, packed and dispatched while you watch; pack standards and insert rules signed off in writing; the exception playbook agreed — who contacts whom, within what time, through which channel; and a first cycle count to establish a stock baseline. A provider that resists any of these steps during courtship, when they are trying to win you, is showing you their standards for later.
One more consideration that surprises brands evaluating for the first time: your sourcing setup changes what a fulfillment partner can do for you. If inbound goods are inspected before they leave the factory — as they should be under any serious sourcing program — your fulfillment inbound becomes verification rather than discovery, and claims for defects land on suppliers instead of dying in a warehouse corner. Evaluate the two together, and for DTC brands scaling past founder-led packing, our DTC brand solutions page shows how sourcing and fulfillment connect as one system rather than two vendors.
Frequently asked questions
How much does a 3PL actually cost?+
Cost is a stack, not a rate: inbound receiving, storage by occupied volume, pick and pack per order, packing materials, shipping by lane, and periodic extras like returns processing or kitting. Modeled against your real order profile, the stack is predictable; quoted as a single blended number, it is not. Insist on line items so you can see which lever is driving your bill as volume grows.
Is one warehouse enough, or do I need a network?+
Most brands are over-served by one well-run warehouse until delivery zones or market expansion say otherwise. A second location only pays once its added stock split and fixed costs are outweighed by faster, cheaper delivery on real volume — which usually happens later than growth excitement suggests. Decide with lane data, not with ambition.
Should my sourcing partner also be my fulfillment partner?+
There is a real operational advantage when one team controls the chain from factory inspection through dispatch: no handoff blind spots, one accountable party for quality, and replenishment planned against actual sell-through. The trade-off is concentration — you depend on one provider's breadth. Evaluate whether they are genuinely strong at both disciplines, not just one with the other bolted on.
When is it too early to move to a 3PL?+
It is rarely about order count and usually about failure signals: packing consumes evenings that product or marketing need, error rates rise with volume, carrier cutoffs get missed, and stock records drift from reality. If none of those exist, self-fulfillment is buying you margin and learning. If they do, the 3PL decision is already late.
