Where Distributors Leak

Problem: You know your revenue to the dollar and your gross margin to one decimal. What you cannot see is that a handful of small, uncontested decisions, a free delivery here, a rush shipment there, a price list nobody has touched in two years, are quietly taking a third of your operating profit.

Quick Win: Do this arithmetic once, today. Take your gross margin (what is left of each sales dollar after you pay for the goods you sold) and your operating profit margin (what is left after you also pay for people, warehouses, and trucks). For US food wholesalers those numbers are 15.44% and 2.61% (NYU Stern, January 2026). Divide one point of gross margin by the operating margin. One percentage point of gross margin is 38% of the entire operating profit. Not one percent. Thirty-eight. That single ratio is why margin leaks in physical distribution behave nothing like margin leaks in a services firm.

Why One Point Of Margin Is Not A Small Number Here

A consulting firm that loses ten hours a week loses ten hours a week. Annoying, recoverable, visible on a timesheet. A distributor that loses one point of gross margin has deleted a large slice of the year, and it will not show up anywhere obvious, because the money never existed as a line item. It leaked out through the cost of the goods you sold before anyone looked at it.

The reason is structural. In distribution, operating costs consume almost the entire gross margin. What is left over is the profit, and what is left over is small.

Here is the arithmetic, using Aswath Damodaran's January 2026 dataset of US public companies. The last two columns are derived: operating costs are gross margin minus operating margin, and the final column is what one point of lost gross margin does to operating profit.

Source: NYU Stern, margins by sector, January 2026. Operating margin is the pre-tax, unadjusted column, and the distributor row is Damodaran's "Retail (Distributors)" category. CSIMarket's wholesale sector data points the same way, with a 12.88% gross margin and a 3.82% operating margin on a trailing twelve-month basis (CSIMarket, Q2 2026).

Read the last column twice. The same leak costs a food wholesaler roughly four times what it costs an industrial distributor and six times what it costs a machinery maker. If you sell food, beverage, or any high-volume low-margin category, you are inside the most unforgiving version of this arithmetic there is.

McKinsey put the flip side of it plainly. Across a database of 130 publicly traded distributors, a 1% price increase was estimated to raise EBITDA margins by 22% (EBITDA is operating profit before interest, tax, and the accounting charge for wear on assets), and to beat a 1% gain in volume, a 1% cut in purchasing cost, or a 1% cut in overhead (McKinsey). Price and margin sit closer to the bottom line than anything else you control. That cuts both ways, which is the point of this post.

Now the five places it goes.

Leak One: The Order That Arrived As A PDF

Somewhere in your building, a person is reading a customer's purchase order off a screen and typing it into your system. Part numbers, quantities, ship-to address, requested date. Maybe it arrived as an email. Maybe a PDF attachment. In some categories, still a fax.

People are good at this and still not good enough. Raymond Panko at the University of Hawaii pulled together a century of human error research and found that on simple mechanical actions, typing a character or a word, accuracy runs 99.5% to 99.8%. On steps that need a thought rather than a keystroke, accuracy falls to 95% to 98% (Panko, EuSpRIG). An order has plenty of both kinds of step in it. APQC benchmarks put the average error rate for manual order entry at 1% to 3% (cited by Conexiom, a vendor, so treat the framing as a vendor's and the benchmark as APQC's).

Take the middle of that band. At a 2% error rate on 40,000 orders a year, that is 800 broken orders.

The labor cost of typing an order is the small part, and every published figure for it comes from a company selling automation software, so treat those numbers with suspicion. The number that matters is the recovery cost of a wrong order, and you can calculate that one yourself: picking it, packing it, shipping it out, shipping it back, putting it away again, issuing the refund, sending the replacement, and very often paying for a rush delivery to make the customer whole. Every one of those lands in the cost of the goods you sold or in freight. Every one comes out of gross margin.

Eight hundred broken orders at even a modest recovery cost is not an efficiency problem. On a 2.6% operating margin, it is a real slice of the year.

Leak Two: Freight You Absorbed And Never Recharged

Most distributors have a free-delivery rule. Orders above some threshold ship free. That threshold was set by someone, at some point, against a freight rate card that no longer exists.

Commonly cited benchmark ranges put total delivery and logistics cost for wholesale and business-to-business distribution at 3% to 8% of sales (a working benchmark range, not audited research). Against a 15% gross margin, delivery cost is somewhere between a fifth and half of everything you make on the goods.

Here is the failure. The threshold does not update itself. Freight rates move, fuel surcharges move, average order size shrinks as customers order more often in smaller quantities, and the threshold sits there unchanged. What was a sensible rule at a $500 average order becomes, three years later, a standing policy of shipping small orders at a loss. Nobody decided that. It drifted.

The test is one query: for last month, plot delivered cost against order value, order by order. If the small end of that chart sits below your gross margin line, you are paying customers to buy from you. This is the physical-goods version of the pattern we mapped in the revenue-leak audit.

Leak Three: Customer Pricing That Drifted Away From Cost

Every distributor has customer-specific price lists. Negotiated once, entered into the system, then left alone for years while purchase costs, freight, and the mix of what those customers buy all moved underneath.

This is the leak with the highest ceiling, because price sits closest to the bottom line. The classic finding from Michael Marn and Robert Rosiello at McKinsey is that a 1% improvement in the price you actually collect lifts operating profit by about 11%, based on the average economics of 2,463 companies (Harvard Business Review, 1992). For distributors specifically, McKinsey's estimate on those 130 public companies was a 22% lift in EBITDA margin per 1% of price (McKinsey).

Two things to look for, both visible in a week:

• Same product, different customers, different prices with no reason. Two customers of similar size and similar order pattern paying materially different net prices is either a deliberate strategy or a leak. Usually it is a leak with a story attached.
• Cost moved, price did not. Sort your top 200 products by change in landed cost (what the goods cost you once freight, duty, and handling are added) over 24 months, and compare it to the change in average selling price to your top 20 customers. The gap is the drift.

We wrote the general version of this in the discount leak. For a distributor the same mechanics apply with a much shorter fuse, because the profit underneath is thinner.

Leak Four: The Rush Shipment That Became Standard

A rush shipment starts as an exception. A customer needed something, you made it happen, everyone felt good. Then it happened again. Then the customer's buyer learned that your standard lead time is negotiable and stopped planning around it.

APQC's benchmarking data, cited by Logility, puts rush freight at 3% of total logistics cost for top performers and 10% for the worst. On an $18M freight budget, closing the gap between a mid-range 7% and a top-performer 3% is worth more than $700,000 a year (Logility, citing APQC). The same analysis notes that 49% of rush events trace back to inaccurate demand forecasts, meaning the company guessed wrong about what customers would order. Half the emergencies are self-inflicted planning failures, not customer chaos.

The premiums are not small. Published freight rate guides put a rush shared-truck shipment at roughly 30% to 60% above standard, a dedicated truck at 50% to 100% above, and air freight at several times the cost of ground (Freight Sidekick, 2026 cost guide). Your own carrier invoices are the better source, and you already have them.

This is a margin leak rather than a cost line because almost nobody recharges it. The rush cost is absorbed to protect the relationship, it goes into freight, and the customer's account still shows the same gross margin percentage it always did, because the rush never got attached to the customer who caused it. Attach it. That one change turns an invisible cost into a conversation you can have.

Leak Five: Stock You Carry For One Customer

Somewhere in your warehouse is inventory that exists because one customer once said they would need it regularly. They stopped pulling it. Nobody removed it.

The standard rule of thumb puts the cost of holding stock at about 25% of its value per year, with expert estimates ranging from 18% to 75% depending on the category, and Helen Richardson's frequently cited breakdown landing at 25% to 55% (Lokad). That charge covers the cost of the money tied up, warehouse space, handling, insurance, taxes, stock going out of date, and stock going missing.

Apply that honestly. $2M of stock held for customers who are not pulling it is roughly $500,000 a year of real cost, and it never appears as a line anyone owns. It sits inside overhead and inside the write-off you eventually take when you admit the stock is worthless.

This is also where the cash side quietly seizes up. We covered the general version in the cash conversion cycle leak, and that post deliberately punted on inventory because it was written for services firms. For a distributor, inventory is the cash conversion cycle. It is the biggest single place your money sits still.

Ranking Five Leaks By What They Actually Cost You

The point of a diagnosis is a ranked list, not a checklist. Here is how the five compare on the two things that decide priority: how big they usually are, and how fast you can size them from data you already have.

Almost every distributor expects the answer to be order entry, because that is the one people complain about daily. It is usually third or fourth. Pricing drift and carried stock are quieter and larger, which is exactly why they survive. The loudest problem is rarely the most expensive one. That is the argument behind a ranked, evidence-backed bottleneck map rather than a list of improvement ideas, and the same logic as the one constraint that limits the whole business.

Where This Breaks: When The Leak Is A Customer You Cannot Fire

Here is the honest failure mode, and it is the reason most of these audits end in a drawer.

You do the work. You attach the true cost to serve each account, meaning everything that customer costs you and not just the goods: freight, rush shipments, returns, refunds, and a carrying charge on stock held for them. You sort customers from most profitable to least. And you find the shape Robert Kaplan and V.G. Narayanan documented at Harvard: the most profitable 20% of customers contribute 150% to 300% of total profit, the middle stretch roughly breaks even, and the least profitable fifth destroys a large share of what is left (Kaplan and Narayanan, 2001). Estimates of the damage at that bottom end vary widely, from 50% to 67% in one common summary (Pragmatic Institute) to considerably worse. Your own number is the only one that matters.

Then you look at the name at the bottom of that list and it is 18% of your revenue.

You cannot fire them. Losing that volume costs you the discounts your suppliers give for buying in bulk, leaves warehouse space empty, and spreads your fixed costs over a smaller base. The analysis was right and the obvious action is wrong. This is where a naive margin audit does real damage.

What works is narrower and slower:

• Recharge the specific behavior, not the relationship. Rush shipments get a documented fee. Below-threshold deliveries get a delivery charge. You are not repricing the customer, you are pricing the exception.
• Change the order pattern before you change the price. Half of the cost to serve on bad accounts is order frequency and order size, not price. A minimum order size often recovers more than a price increase and starts a much easier conversation.
• Fix the leaks that need no customer conversation first. Order entry errors, carried stock, and rush shipments you caused yourself are all yours to fix alone. Do those before you touch a single account.

The other failure mode is subtler: sizing the leaks and then never assigning one. A margin leak with no owner is a slide, not a fix. If nobody's name sits next to "delivery threshold," the threshold does not move.

What A Real Diagnosis Produces

Not a benchmark deck comparing you to a peer group. A ranked list built from your own order lines, freight invoices, price lists, and stock movement, where every item carries the evidence that proves it and a number you can defend in a room. At a 2.61% operating margin, you do not have twelve months to find out.

If you want a ranked map of where your margin is actually going, built from your own data rather than an industry average, that is what we install.