Distribution costs are every expense required to put a finished product on the market and deliver it to the customer · not just freight. For a Mexico–US shipper, that mixes selling, DC/CEDIS, transport, last mile, shrinkage, and friction (detention-style wait, accessorials, distribution-center chargebacks).
This guide separates commercial cost from delivery logistics cost, compares distribution vs landed cost vs freight spend, shows cost-to-serve by customer/corridor/channel, and ties the math to operable evidence (TMS, GPS, POD, CFDI) plus a worked example in Mexican pesos.
- commercial vs delivery logistics
- 2 layers
- distribution · landed · freight
- 3 metrics
- typical recovery auditing 100% of freight
- 5–7%
- measured pilot without a TMS big bang
- 6–8 wks
Cluster context: 7 money leaks in freight · CFDI + Carta Porte + GPS + POD audit · OCL calculators.
What distribution costs are (short answer)
They are the set of expenses and investments required to move a finished product to the point where it generates a sale: channel, customer DC, or end consumer. That includes commercialization and delivery logistics.
What they are not: only the freight line on the P&L; only “ads + commissions”; or a classroom allocation exercise disconnected from the yard.
In Mexico, a useful market reference is to express distribution as a percent of net sales (warehouse + packaging + transport). That metric only helps if you also capture friction: wait, redeliveries, deductions, and accounts payable work.
Two layers: commercial vs delivery logistics
The distinction that changes decisions: separate what it costs to sell from what it costs to deliver with evidence. Thin SEO posts often list categories and jump to routing; here the budget owner is different.
Distribution · two layers
Selling vs delivering with evidence
Same P&L row · two owners and two levers.
Commercial layer
Cost to sell
Commissions, offices, trade marketing · credit/collections if booked here.
Logistics layer
Cost to deliver
CEDIS, freight, last mile, shrinkage, detention, accessorials, and chargebacks.
Rule
If you mix commissions with detention, you are not measuring distribution · you are averaging leakage.
If finance rolls everything into “distribution expense,” operations cannot attack detention and accounts payable cannot hold an invented accessorial. Split the layers even when the financial statement presents them together.
Distribution vs landed cost vs freight spend
Three numbers that get confused in leadership meetings. If you mix them, “cut distribution costs” becomes “cut the spot rate” · and leakage stays untouched.
Freight spend
What it includes: Rate + billed tolls/FSC + accessorials paid to carriers
What it is for: Negotiate carriers and audit invoices
Typical mistake: Believing it is the full cost to deliver
Distribution cost
What it includes: Commercial + delivery logistics (warehouse, last mile, shrinkage, delivery admin)
What it is for: Price, channel, network mix
Typical mistake: Booking only freight or only commissions
Landed cost
What it includes: Product cost + transport + duties/insurance/handling to the node
What it is for: Compare origins / modes / suppliers
Typical mistake: Using it as a synonym for commercial distribution
On the Mexico–US corridor, freight spend is usually the most visible and the easiest to “optimize” blind. Distribution cost explains margin by customer. Landed cost matters when you choose origin, mode, or consolidation · not when accounts payable releases a tractor payment.
Useful classification vs what operations loses
University notes (object / function, allocation) are not wrong · they are incomplete. The academic lens says what was spent; the operable lens says where pesos leak after the “all-in” quote.
By object (academic)
What it answers well: Labor, freight, warehouse, packaging, shrinkage, systems
What it usually omits (real leakage): Accessorials, detention, redelivery, POD dispute, wrong rate
By function · product
What it answers well: Which SKU does not pay for its delivery
What it usually omits (real leakage): Allocating by volume when freight is priced per trip
By function · territory
What it answers well: Which corridor is expensive (e.g. Bajío–Laredo)
What it usually omits (real leakage): Tolls, security, border wait time
By function · customer / channel
What it answers well: Who is expensive to serve; whether the channel absorbs it
What it usually omits (real leakage): Punitive DCs, chargebacks, impossible windows
Allocation still exists (not everything is direct). Modern rule: allocate only what you cannot attribute with evidence (trip, appointment, POD), and document the driver (km, pallets, orders, dock hours).
Why they can exceed production cost
In the plant, unit cost usually stabilizes with volume. In distribution, every delivery restarts friction: appointment, wait, partial refusal, documentation, freight collections. That is why getting product to the shelf can cost more than making it · especially with a fragmented network and sampling in accounts payable.
- Distance and mode: more nodes = more handoffs and damage risk.
- Service level: tight windows and OTIF (on time in full) buy expensive capacity. Related: OTD (on-time delivery).
- Powerful customers: DC deductions and chargebacks push cost back to the shipper.
- Administration: tower + accounts payable without a trip file is fixed labor that scales with volume.
If your dashboard only watches manufacturing cost and list price, margin evaporates in distribution with no owner.
Cost-to-serve by customer, corridor, and channel
Cost-to-serve is the total cost attributable to serving a unit of analysis · customer, corridor, or channel · including visible freight and friction. It is not the company-wide average.
First freight and operating friction; then tower/accounts payable admin and the decision signal per unit.
Customer A · local wholesale
Base freight: $1,200
Friction: $80
% friction: 7%
Customer B · retail DC
Base freight: $1,450
Friction: $620
% friction: 43%
Bajío–Laredo corridor
Base freight: $28,000
Friction: $4,200
% friction: 15%
E-commerce channel
Base freight: $95
Friction: $70
% friction: 74%
Customer A · local wholesale
AP admin: $40
Cost-to-serve: $1,320
Signal: Healthy if price covers
Customer B · retail DC
AP admin: $180
Cost-to-serve: $2,250
Signal: Review SLA and deductions
Bajío–Laredo corridor
AP admin: $900
Cost-to-serve: Per trip
Signal: Audit 100% of the flow
E-commerce channel
AP admin: $35
Cost-to-serve: $200
Signal: Does the ticket absorb it?
Use both tables for three decisions: raise price or a service fee; change mode/window; or stop subsidizing the expensive customer with the easy one’s margin. For per-delivery/km detail, see cost per delivery and per km.
Hidden leaks: not “freight only”
This is where you beat guides that list “sales + warehouse + transport + shrinkage” and stop. On Mexico–US lanes, delivery logistics cost inflates with leaks that never appear on the “all-in” quote.
Detention / wait time
How it shows in pesos: Extra hours at dock or yard
Minimum evidence: Appointment + GPS/geofence + timed POD
OCL guide: Detention
Accessorials and false freight
How it shows in pesos: Handling, dry runs, uncontracted layover
Minimum evidence: Rate confirmation + invoice
OCL guide: Accessorial calculator
Failed delivery / redelivery
How it shows in pesos: Second trip or penalty
Minimum evidence: POD + refusal reason
OCL guide: POD convention
POD disputes
How it shows in pesos: Chargeback or customer payment hold
Minimum evidence: POD convention + photos/signature
OCL guide: Proof of delivery
Invoice / CFDI errors
How it shows in pesos: Overpay or pay twice
Minimum evidence: Rate + CFDI + Carta Porte + GPS + POD
OCL guide: Invoice audit
DC deductions (food/retail)
How it shows in pesos: Post-delivery discount
Minimum evidence: Delivery vs order reconciliation
OCL guide: Processed foods
Go deeper on 7 money leaks, detention, POD convention and routing (when the leak is distance/stops, not only the rate).
How to calculate: formulas + process
Start with simple formulas. Then attribute. Then audit freight · not the other way around. Calculators speed the scenario; they do not replace the trip file.
CTS = Freight + Accessorials + Attrib. warehouse + Shrinkage + Delivery admin
- CD
- Distribution cost for the period (commercial + delivery logistics).
- VN
- Net sales for the same period (after relevant returns).
- CTS
- Cost-to-serve for the unit (customer / corridor / channel).
Operable calc
From data to cost
Freeze period
Month and corridor
Split layers
Commercial vs delivery
Sum objects
Freight and warehouse
Attribute function
Customer or channel
Audit freight
100% of pilot
Useful tools while you build the model: freight calculator · accessorials · audit ROI · detention. The TMS records trips and rates; it does not replace the cross-check with CFDI, Carta Porte, GPS, and POD. See TMS guide for Mexico.
Worked example (MXN · corridor + DC)
Grocery shipper.One month. Regional Bajío corridor to two chains (one with a strict DC). Rounded figures to teach allocation logic · not a classroom “chocolates and gummies” case.
Sales commissions
Amount (MXN): $180,000
Layer: Commercial
Allocation driver: By sales (already in CRM)
Warehouse / own DC picking
Amount (MXN): $220,000
Layer: Delivery logistics
Allocation driver: Orders / lines
Contracted freight (invoiced)
Amount (MXN): $410,000
Layer: Delivery logistics
Allocation driver: Trip / rate
Detention + accessorials paid
Amount (MXN): $48,000
Layer: Delivery logistics
Allocation driver: Trip with evidence
In-transit shrinkage
Amount (MXN): $22,000
Layer: Delivery logistics
Allocation driver: SKU / trip
Tower + accounts payable hours
Amount (MXN): $35,000
Layer: Delivery logistics
Allocation driver: Corridor trips
Total CD for the month
Amount (MXN): $915,000
Layer: —
Allocation driver: —
Customer split: Chain B concentrated $31,000 of detention/accessorials and $12,000 of POD dispute hours. Its cost-to-serve per order landed ~40% above Chain A, even though base freight per km was similar. Decision: window fee + pre-pay audit on that corridor · not “lower average freight” blind.
Data-source checklist
Without sources, the calculation is opinion. Cross at least these pieces before reporting cost-to-serve to leadership.
Elige un paso para ver el detalle
Detalle del paso · 01
TMS / trip orders
Objectives for management decisions
Measuring distribution is not an accounting end in itself. It is for deciding. Five concrete uses:
- Price and fees: move the cost of impossible windows to the customer who demands them.
- Channel mix: stop subsidizing e-commerce or export with wholesale margin.
- Network and mode: consolidate, change DCs, or split expensive corridors · see also routing.
- Carrier negotiation: talk with detention and accessorial evidence, not vibes.
- Freight payment: hold what does not match; pay what is verifiable.
If the indicator does not change one of those five levers in 90 days, you are reporting · not managing. Fleet preventive maintenance does not replace this calculation, but it prevents roadside leakage; see preventive maintenance limits.
What OCL executes before you pay
OCL Cargo is an autonomous TMS with computer use agents: the agent operates interfaces (portals, email, PDFs) the way your tower would, and leaves a trip file · not another screen to type into by hand.
In distribution, the direct impact is on the delivery logistics layer: audit before you pay by crossing rate, CFDI, Carta Porte, GPS, and POD. Moving from sampling to 100% of the pilot flow, the published pattern is recovering 5–7% of audited spend. Typical pilot: 6–8 weeks · on the order of ~$50 MXN per shipment.
It coexists with your TMS or spreadsheet: no day-one migration. Who decides exceptions? Your team. The agent closes the repeatable work. OCL can stamp invoice and Carta Porte.
Before payment
A trip file that matches
Take the trip
From your TMS
Cross docs
CFDI and Carta Porte
Validate route
GPS and geofence
Tie POD
Clear convention
Pay or hold
With trip file
More context: computer use in logistics · false digitization.
Key takeaways5 points
- Distribution cost ≠ freight only: it includes commercial (sales/promo) and delivery logistics (warehouse, transport, last mile, shrinkage, detention, accessorials, DC deductions).
- Separate three metrics: freight spend · distribution cost · landed cost. Mixing them hides leakage.
- Classify by object and by function · without that, allocation lies. Academic lenses help; operable friction is what pays (or kills) margin.
- Cost-to-serve by customer/corridor/channel explains why the “same” price destroys margin on one side and looks fine on the other.
- Auditing 100% of pilot freight typically recovers 5–7% of audited spend; OCL coexists with your TMS · about $50 MXN/shipment · your team decides exceptions.
Is your distribution % hiding freight leaks?
Related reading
Frequently asked questions
They are every expense required to put a finished product on the market and deliver it to the customer (or channel), including what it costs to sell and to deliver with evidence · not just the line-haul freight invoice.
No. Freight spend is what you pay carriers. Distribution cost adds commercial + delivery logistics (warehouse, last mile, shrinkage, detention, accessorials, deductions). Landed cost is the full cost of placing an SKU at a node: product + transport + duties/insurance when applicable. Mixing them hides where margin leaks.
Because the unit left the plant with a relatively stable manufacturing cost, while delivery restarts friction every trip: warehouse, miles, wait time, refusals, chargebacks, and administrative work. On Mexico–US corridors, operating friction often grows faster than manufacturing cost.
It is the total cost of serving a specific customer, corridor, or channel · not the company average. Two customers with the same list price can have opposite margins if one drives detention, redeliveries, and proof-of-delivery (POD) disputes.
Freeze a period (month) and one corridor. Add freight + accessorials + attributable warehousing + shrinkage + tower/accounts payable (AP) hours. Cross transportation management system (TMS), GPS, POD, and Mexico’s digital tax invoice (CFDI). Use this article’s data-source checklist and the OCL calculators.
A large share of delivery logistics cost is paid blind. When you audit 100% of the pilot flow (rate + CFDI + Carta Porte + GPS + POD), the typical pattern is recovering 5–7% of audited spend. See freight invoice audit and the audit ROI calculator.
Not on day one. OCL is an autonomous TMS with agents (computer use: the agent operates screens like your control tower): it builds the trip file on top of or beside your system of record. Typical pilot: 6–8 weeks · about $50 MXN per shipment. Who decides exceptions? Your team. OCL can stamp invoice and Carta Porte.