Warehouse cost analysis, 3PL prices and automation

The pricing of 3PLs can be complicated, but does it need to be?

The answer is that if you want to be "100% fair", you need a complicated pricing structure. But what many companies have realized, is that people hate complicated pricing structures, for example people would rather want to pay a premium for a fixed price of data per month instead of paying per MB, even if it was cheaper paying for actual usage.

We will here look at how we can simplify the pricing of 3PLs by baking the different costs inside the price per orderline, so that we only charge per orderline. 

We look here at a real 3PL offer for a company doing about 200-300 orders/day with an average 2 orderlines per order. 

Prismodell Kost/enhet
Per ordre/retur kr 18
Varelinje kr 5
Pr stk. Produkt kr 1
Timepris varemottak kr 500
Timepris adhoc kr 500
Lagerleie hylle kr 20
Lagerleie pall kr 85
Systemkostnad kr 990
Informasjon Estimat
Antall ordre/mnd 15000
Antall unike varelinjer per ordre 2
Antall stk. Produkter per varelinje 2
Antall unike produkter på lager 3000
Antall per unike produkt/gj. 10
Totalt antall produkter på lager 30000
Antall varemottak/mnd 10
Snitt antall unike varelinjer/varemottak 30
Antall unike varelinjer varemottak/måned 300
Antall stk per varelinje (10 skt) 3000
Antall returer per måned 10 % 1500
Lagerdrift Antall Kostnad Total kostnad
Utsendelse av ordre 15000 kr 32 kr 480,000
Varemottak 20 kr 500 kr 10,000
Returhåndtering 1500 kr 32 kr 48,000
System 1 kr 990 kr 990
Lagerleie hylle (85%) 510 kr 20 kr 10,200
Lagerleie pall (15%) 225 kr 85 kr 19,125
Totalt kr 568,315
Kostnad per ordre (lagerdrift) kr 38
Frakt % fordeling Antall Kostnad Total kostnad
Helthjem / Bring pakke i postkassen (0-2 kg) 85% 12750 kr 30 kr 382,500
Bring utlevering i butikk (Gjennomsnittvekt <35 kg) 10% 1500 kr 60 kr 90,000
Ekspress/ levering samme kveld 5% 750 kr 100 kr 75,000
Totalt kr 547,500
Fraktkostnad per ordre kr 37
Total kostnad lagerdrift og frakt kr 1,115,815
Gjennomsnittlig kostnad per ordre kr 74.4



Cost "merging"

Looking at this from a high level, we could say that other than carrier costs which are invoiced without surcharge, the total cost is 568815 kr per month. 

Instead of 568815 kr / 30000 = 19 kr, or 38 kr per order as the table above shows. 

So one could argue it would be better to just charge 19 kr per orderline and then ignore the complexity of invoicing for everything separately. Keeping track of invoicing itself has a non-negligible cost. 

The question then is if only charging per orderline is unsustainable or less profitable. What if the company is not selling anything? What if they keep too much stock? What if they have an abnormal amount of returns? What if the returns are complicated and the customer starts asking for special packaging instructions etc.

What might work is to bake uncertainty like this into the price and then also having in the terms of service a max amount of stock before charging extra. But that again adds the complexity of making functionality and routines for charging per storage unit. Then one could argue to just charge per storage unit anyways. 

There are obvious risks here if the return percent is 20%, the pricing model would give 20.5 per orderline. 

 To evaluate how much risk to take by having a simpler pricing structure, let's look at the profits, let's look at the numbers:


5 kr per orderline (automation cost) + 3.7 kr pick cost  + 0.5 kr emballasje+label 

= 9.2 per orderline

pick cost is calculcated: (190*7.5*1.3)/500, given that a warehouse worker picks 500 orderlines in a 7.5 hour shift and has an hourly wage of 190 kr. Depending upon the efficiency of the warehouse operations, this might be more or less. The 30% factor is used to factor in taxes and extra costs that comes with employment.

That means per orderline you will earn about 9 kr charging the above.

This easily covers for a normal return rate, but keep in mind that reverse logistics is more complicated and that handling of it will cost more in terms of warehouse worker time. 

What I propose is that the pricing assumes a high return rate, and that customers that prove a lower return rate, might get a lower rate for example. This gives a good incentive to reduce something you want to reduce anyways. 

For storage, in one way, the incentives are aligned: No one wants to have a lot of stock. But that does not stop business owners from having overstock from time to time. By taking a more active role in advising customers in purchase management / order planning / replenishment analysis, one could together save costs and keep pricing simple. But if there is too much uncertainty and you think the customer will abuse their "free" storage, charging for storage might be necessary. 

With regards to other costs like termination costs etc. It is probably best to not surcharge any for this to remain on good terms. Business might boom in the future. 

Finishing off I would say that maintaining a very simple price structure might be possible if:

  • You do due diligence 
  • You have good communication with the customer, to minimize cost for both sides
  • You factor in the risk of high returns and abuse of storage into the price

 The advantages of simpler prices:

  • Less cost invoicing, no more discussions over "this extra cost"
  • Less complex systems
  • More predictable pricing for customers
  • Higher conversion rate on sales? (easier to choose you as 3PL)


 Automation cost analysis

First, what is the running cost of running your own warehouse:

  1. Employees
  2. EDI/software
  3. Lease

 The employee cost and lease will vary a lot from country to country. But in Norway you can expect to pay warehouse workers about 16-25 euro usually at a "professional" warehouse. Other countries can pay much less, but you will seldom see less than 10 euro for most western countries. You can usually multiply the hourly rate by 1.3 to get the total cost of employees as "employee tax" and other taxes usually apply. 

The lease is country, city and location dependent and also the standard of the building is a factor. 

 For Oslo in Norway, about 100 euro+ per square meter per year is not unusual for a warehouse facility. So if you have 1000 m2, then you pay 100 000 euro a year. 

Let's say you need 5 people to run this warehouse with an average cost (including taxes etc.) of 25 euro/hr per person.

Then the total monthly cost is, assuming only working on regular days: 25*7.5*20 = 3750 euro, per person.

So let's look at the cost savings you get from an automated autostore/pio warehouse.

Depending upon the height of your warehouse, the amount saved will be different, but a good minimum assumption is at least 20% cut in need for space, so instead of 1000 m2, you need 800 m2. 

That means you save 20k euro each year at least on the space. 

With regards to employees, it is hard to calculate, but based on experience from fulfillment for famme.no, the total personell on warehouse+customer service should be halved. 

Let's look at the cost, assuming 200 orders/day on average:

350 orderlines = 200 orders (1.7 average).
350*0,5 = 175 euro. (pio cost 0.5 per orderline)

175*20 = 3500

This is what you need to save in human labor monthly to be "break even".

So say that you save 2 persons which cost is 25 euro. You have to take customer service into account as well when counting persons saved, so saving two persons in total is a mild estimate depending on how efficient the warehouse was from the get go.

That is monthly savings of:

2*3750 = 7500 euro

Some other savings which are harder to estimate:


  • added control over stock.
  • Stock counting related to auditing  
  • less loss of stock
  • items that are in the warehouse but not in online store

As we can see, the savings can be a lot:

7500-3500(pio cost) + 20/12 = 5666 euro in monthly savings. 

Let's say that the accuracy of picking increases so that we avoid 3 mistakes/day. A mistake can be wrong item or missing item. We assume that the total cost of all mistakes is at least 100 euro (it could be more in lost customer value). 

100*20 = 2000 euro monthly in savings due to higher efficiency and accuracy. 

Monthly savings is now:

5666+2000 = 7666 euro

Let's round these savings up to 8000, accounting for less loss of stock, faster shipping to customers (increased customer value) etc:

8000*12 = 96k euro in total yearly savings. 

Let's assume the difference in capex between an automated warehouse and a normal warehouse with racks is 150000 euro, then:

150000/96000 = 1.6 years to save in the capex. 

Given that the numbers here are correct, the time before ROI is very little. 

So the only question for these countries is when do you expect to get ROI on the invested capex.

I strongly believe these numbers are not even exaggerated, as things like learning curve for pickers etc. are drastically lowered with an automated system. There are of course also some costs and risks that increases with automation, so I think all in all the math above is "fair" to both sides.  


Back to blog