Logistics Questions to Ask
Questions for anyone getting up to speed on a logistics operation, whether you are new to the team, evaluating a provider or reviewing a supply chain. They cover the numbers that get watched daily, inventory accuracy, cycle time, cost per order, carrier concentration, peak staffing, damage, customs delays and what the operation would fix first.
The questions
Open any question for the note
Which numbers do you look at every morning, and which one tells you first when something is wrong?
Why ask it
The daily list reveals what the operation actually manages, as opposed to what appears in the quarterly deck. If the answer is a long dashboard nobody names a favorite from, alerting is probably done by phone call from an angry customer. A single leading number, such as orders not yet picked by ten o'clock, is a sign of a mature operation.
What is your on-time delivery rate, and how do you define on time?
Why ask it
The definition is where the number is made. On time against the customer's requested date is a much harder measure than on time against the date the operation promised after it saw the order. Ask which one is being quoted, and whether a delivery moved by agreement still counts as on time.
How accurate is your inventory record against what is actually on the shelf?
Why ask it
Anyone who says the record is perfect has not counted recently. What you want is the cycle count variance and how often it is measured, because inventory error propagates into every other number: phantom stock creates promises the warehouse cannot keep, and hidden stock creates purchases nobody needed.
How long does an order take from the moment it lands to the moment it ships, and where does most of that time go?
Why ask it
Total cycle time is easy to quote and rarely broken down. The interesting part is almost always waiting rather than working: an order sitting for six hours until the next wave releases, or a credit hold nobody watches. Ask for the split and you find the fixable hours.
Where do shipments most often get stuck?
Why ask it
Everyone on the floor knows the answer to this and it seldom matches the process map. Expect a specific dock door, one supplier's paperwork, or the handoff between two systems. If the answer is a shrug, ask to see the last twenty late orders instead.
What does it cost you to move a typical order, and what is included in that figure?
Why ask it
Cost per order is only comparable when you know whether it includes labor, packaging, returns and the warehouse itself. A figure that covers freight alone will look impressively low and hide where the money goes. Watch for accessorial charges, which often add a fifth to the freight line and appear nowhere in the quoted rate.
Which of your logistics costs are fixed, and which move with volume?
Why ask it
This determines what happens in a downturn and how much a growth plan will really cost. Operations with heavily fixed cost bases can be efficient at volume and dangerous when demand drops. If nobody can split the two, cost modelling for any change will be guesswork.
What happens on your side when a carrier misses a pickup?
Why ask it
You are asking about the exception path, which is where service is actually won or lost. Good answers describe who notices, within what window, and who calls the customer. Poor answers describe finding out the next day from a tracking report.
How did you choose your current carriers, and when did you last test the market?
Why ask it
Rates and service both drift, and incumbency is comfortable. A relationship that has never been benchmarked usually costs more than it should, but frequent switching has its own price in learning curves and damaged goods. Look for a deliberate cycle rather than either extreme.
How much of your volume moves through a single carrier, lane, or warehouse?
Why ask it
Concentration is the risk that does not show in the numbers until it does. Ask for the percentage through the largest node, then ask what the plan is if that node is unavailable for a week. Many operations discover they have one supplier for a component they thought was dual sourced.
If your main warehouse were unusable tomorrow, what would you do in the first forty-eight hours?
Why ask it
The value is in whether the answer is specific: a named alternative site, an agreement already signed, a list of what would ship first. A plan that exists only as a document nobody has rehearsed is worth knowing about now rather than during a flood.
What does your peak week look like compared with an average week, and how do you staff for it?
Why ask it
The ratio tells you how much of the year the operation is designed for. A four to one peak run on temporary labor means quality and safety both dip when volume rises, and that is when new customers form their first impression. Ask what error rates do during peak, not just throughput.
When an order is going to be late, how does the customer find out?
Why ask it
Separates real visibility from a tracking page. The good answer is that the system flags it and someone contacts the customer before the promised date passes. The common answer is that the customer calls to ask, which is the single most expensive way to handle a delay.
What do customers complain about most, and what causes it?
Why ask it
The complaint and the cause are often in different departments, which is why asking both halves matters. Damage blamed on carriers frequently traces back to packaging decisions made to save a few cents per unit. Whoever answers this will also show you how blame moves through the organization.
How much of this work still happens in spreadsheets, email or over the phone?
Why ask it
Manual steps are where errors and key-person risk live, and they rarely appear in a systems diagram. Ask who maintains the spreadsheet that everything depends on, and what happens during that person's time off. The answer is usually one name.
What did your last system or process change actually cost you in disruption?
Why ask it
This tests institutional honesty and tells you how the next change will go. Operations that can describe the messy first month, the temporary workarounds and what they underestimated are safe to change again. Operations that describe a flawless rollout are either lucky or not telling you.
How much of your damage and loss can you trace to a specific step?
Why ask it
Untraceable shrink means the handoffs are not instrumented, and that gap costs more than the damage itself because nothing can be fixed. Ask whether damage is recorded where it is found or where it is thought to have happened, since those are different data sets.
Where do you hold safety stock, and how were those levels set?
Why ask it
Levels set years ago from a supplier lead time that has since changed are extremely common. Listen for whether the calculation accounts for demand variability or only average demand, because the second produces stockouts precisely when demand spikes.
For your international freight, what has held up customs clearance in the past year?
Why ask it
Border delays repeat, because the causes are usually documentary rather than random: a classification dispute, a missing certificate, one broker who is slow. Specific past incidents predict future ones far better than a general statement that clearance is fine.
If you could fix one thing here and money were not the constraint, what would it be?
Why ask it
Asked last, once trust has built, this is where you learn what the person closest to the work believes. It often names something structural that no metric captures, and the gap between this answer and the official improvement plan is the most useful thing you will take away.
Getting Straight Answers About an Operation
Practical guidance for the conversation itself
Check the Answers Against Records
Ask for the exception report, not the summary
Averages hide the cases that generate complaints and cost. Ask to see the last twenty late orders, the last ten damage claims or the last month of stockouts. Patterns show up in the tail long before they move a monthly average.
Walk the floor before you accept the process map
Spend an hour at the receiving dock and the pack bench and watch what people do with their hands. Workarounds, second keyboards, handwritten notes taped to a monitor and pallets in the wrong aisle each tell you where the documented process stopped matching reality.
Ask the same question at two levels
Put the question about where shipments get stuck to a manager and to a picker on the same day. Where the answers differ you have found either a reporting gap or a problem nobody has been rewarded for raising.
Reading the Numbers You Are Given
Always ask for the definition and the denominator
On-time rates, fill rates and cost per order can each be calculated several defensible ways, and the flattering version is usually the one presented. Ask what counts as a failure and which orders were excluded before you compare anything to a benchmark.
Ask what the number was twelve months ago
A direction matters more than a level, and it is harder to dress up. If nobody can produce the prior year figure, the metric is probably assembled for meetings rather than used for decisions.
Convert percentages into orders and dollars
Ninety-six percent on time sounds strong until you translate it into four hundred unhappy customers a month. Doing the arithmetic out loud in the room changes the conversation faster than any argument about targets.
What Goes Wrong in These Conversations
- Asking about best practices instead of about this operation. General questions get general answers and teach you nothing about the site in front of you.
- Accepting technology as an answer to a process question. A new system laid over an unclear handoff produces faster confusion.
- Talking only to the people who present well. The shift supervisor and the returns clerk know things the operations director has never been told.
- Treating a single quarter as a trend, particularly one that included a peak or a system migration.
- Failing to ask who is not in the room. If the carrier, the supplier or the customer service team is absent, at least half of the delay you are investigating happens where you cannot see it.
- Letting the visit end without asking what the person would change. It is the one question that surfaces problems nobody has been able to escalate.