# How We Save You Money | AutoScheduler.AI

> Canonical page: https://autoscheduler.ai/savings/
> "AI" doesn't magically mean savings. This page shows exactly how AutoScheduler's Warehouse AI turns operational improvements into dollars: 21 savings levers, the math behind each one, published customer proof, and a calculator a site can run with its own numbers.

AutoScheduler's position: a percentage improvement only matters once it's translated into dollars. Every lever below shows the equation, where the starting numbers come from, and how to check the result against your own WMS, yard, time clock, or finance data.

- [Request a free site visit](https://autoscheduler.ai/request-a-site-visit/): we walk your floor and size the prize
- [Book a demo](https://autoscheduler.ai/book-a-demo/): full platform walkthrough with our team

## How the calculator works

1. **Set your rates.** Floor pay with benefits (default $28/hr), office pay with benefits ($35/hr), overtime pay ($42/hr), days open a year (300), forklift speed (4 mph), and forklift cost per mile ($1).
2. **Pick what costs you money.** Each lever you pick opens its own calculator.
3. **Put in your numbers.** The page totals estimated savings a year and splits them into four kinds: sales you keep, fees you stop paying, hours you don't work, and room to grow. Hours turn into cash when you cut overtime, use fewer temps, or leave open jobs open.

Units per hour (UPH) already includes travel, pick density, and overtime, so the page warns you not to count both and double-count the same savings.

## Prove it: measure it one load at a time

1. **Agree on rates first.** Finance signs off on every rate before anyone measures.
2. **Compare plan to no plan.** The gap between loads that followed the plan and loads that didn't is the savings.
3. **Count only real work.** Only hours spent moving product count.
4. **Leave out surges.** Overtime from a volume spike is fair; it isn't counted as waste.

Example: **$292.87 saved on one 24-pallet trailer** by following the plan, using the default rates:

| Item | Saved |
| --- | --- |
| 45 cases not cut × 35% lost sales × $20 × 30% margin | $94.50 |
| 2.5 hours of detention avoided × $75 | $187.50 |
| 0.75 forklift miles avoided: 0.19 hours × $28 + 0.75 miles × $1 | $5.97 |
| 0.35 overtime hours moved to regular time × ($42 − $28) | $4.90 |

## How a project works: pick one problem to attack together

1. **Pick one problem:** the biggest number in your calculator.
2. **Agree on the number:** same math, your data.
3. **Automate it:** plans land in your WMS.
4. **Measure it:** every shift, plan vs. actual.

## The 21 savings levers

Proof points name the customer where results are public and use a descriptor (for example "Top 5 global F&B manufacturer") where they aren't. Worked examples use the calculator's default inputs and are illustrations, not customer results.

## Sales and service

### Fill rate failures

*Sales you keep.* Each point of fill rate is cases that ship instead of getting cut.

**Formula:** Cases × points gained × lost-sale rate × price × margin

**Proof:**

- −50% controllable cuts (CPG case study, https://autoscheduler.ai/wp-content/uploads/2025/01/ASI_Gen_CS-Final-11-7-24-2.pdf)
- >99% case fill held (Consumer foods case study, https://autoscheduler.ai/wp-content/uploads/2023/08/AUT_CaseStudy_EnhancedAutoPilot-LowRes.pdf)
- +2% increase in fill rate (Fortune 100 F&B)

**The math:**

- Cases saved = cases ordered × fill rate gain
- Lost sales avoided = cases saved × lost-sale rate
- Sales kept = lost sales avoided × price per case
- Margin kept = sales kept × margin

**Illustrative example (the calculator's default inputs, not a customer result):**

- 10,000,000 cases × 1 points = 100,000 more cases shipped
- 100,000 × 35% = 35,000 cases that would have been lost sales
- 35,000 × $20 = $700,000 in sales kept
- $700,000 × 30% = $210,000 in margin kept

**Where the starting numbers come from:** FMI, GMA and CIES out-of-stock study (2002). When an item is out of stock, 31% of shoppers buy it at another store, 26% switch brands and 9% buy nothing. For a maker, brand switches and no-buys are lost sales (35%). For a retailer, other-store buys and no-buys are lost (40%).

**Check it yourself:** Pull cases ordered and cases shipped by day from your WMS, with the cut reason. Finance has your price and margin per case.

### Retailer fines (OTIF)

*Fees you stop paying.* Retailers fine loads that show up late or short.

**Formula:** Goods on late loads to retailers that fine × fine rate × share you control × cut

**Proof:**

- 50→80% loads ready on time (CPG plant on SAP EWM)
- $2–3M a year across 10 plants (Our model)

**The math:**

- Goods on missed loads = loads × share to retailers that fine × goods cost per load × share that miss
- Fines = goods on missed loads × fine rate
- Fines you can fix = fines × share caused at the dock × share actually charged
- Saved = fines you can fix × cut

**Illustrative example (the calculator's default inputs, not a customer result):**

- 12,500 loads × 50% × $25,000 × 12% = $18,750,000 in goods on missed loads
- $18,750,000 × 3% = $562,500 in fines
- $562,500 × 60% × 75% = $253,125 you can fix
- $253,125 × 40% = $101,250 saved

**Where the starting numbers come from:** 8th & Walton, Walmart OTIF. Walmart fines 3% of the cost of goods on cases that miss. Prepaid suppliers must hit 90% on time and 95% in full.

**Check it yourself:** Get retailer scorecards and fine reports from finance. Match them to load ship times in the WMS. That shows which misses started at the dock.


## Dock and yard

### Live detention

*Fees you stop paying.* Carriers bill by the hour when a live load runs past free time.

**Formula:** Billed hours × hourly rate × cut

**Proof:**

- −33% detention and dwell (CPG case study, https://autoscheduler.ai/wp-content/uploads/2025/01/ASI_Gen_CS-Final-11-7-24-2.pdf)
- 30–50% lower detention (P&G, PepsiCo)
- $171K scoped at one dc (Manufacturer)

**The math:**

- Billed hours = live loads × share billed × average billed hours
- Detention cost = billed hours × hourly rate (or use your real bill)
- Saved = detention cost × cut

**Illustrative example (the calculator's default inputs, not a customer result):**

- 20,000 loads × 15% × 1.5 hours = 4,500 billed hours
- 4,500 hours × $75 = $337,500 in detention a year
- $337,500 × 35% = $118,125 saved

**Where the starting numbers come from:** ATRI, Costs and Consequences of Truck Driver Detention (2024). 39.3% of stops had detention in 2023. 67% of carriers bill after 2 hours. Median rates are $50 an hour for dry vans, $60 for refrigerated and $80 for special freight. Carriers bill 75% of detention time and get paid on 55% of those bills.

**Check it yourself:** Use carrier detention bills. Or compare appointment times to gate-in and gate-out times in your yard system.

### Dropped trailer fees

*Fees you stop paying.* Carriers charge by the day when a dropped trailer sits too long.

**Formula:** Days past free time × daily fee × cut

**Proof:**

- site-built app ranks trailers by the shortages they cover (Top 5 global F&B manufacturer)

**The math:**

- Extra days = trailers dropped × average days past free time
- Fee cost = extra days × daily fee
- Saved = fee cost × cut

**Illustrative example (the calculator's default inputs, not a customer result):**

- 6,000 trailers × 0.8 days = 4,800 extra days
- 4,800 days × $50 = $240,000 a year
- $240,000 × 30% = $72,000 saved

**Where the starting numbers come from:** FreightWaves trailer fee research. The average daily fee is $61.70. The median is $64 and the most common is $50. Median free time is about 2 days.

**Check it yourself:** Use yard check-in dates and empty-release dates for each trailer. Compare them to each carrier's free time.

### Campus shuttles

*Fees you stop paying.* Product shuttled out and back gets paid for twice.

**Formula:** Trips that come back × cost per trip × cut

**Proof:**

- −45.6% bring-backs (P&G)
- +46.3% loads shipped straight from the plant (P&G)

**The math:**

- Bring-backs = shuttle trips × share that come back
- Cost = bring-backs × cost per round trip
- Saved = cost × cut

**Illustrative example (the calculator's default inputs, not a customer result):**

- 25,000 trips × 10% = 2,500 bring-backs
- 2,500 × $125 = $312,500 a year
- $312,500 × 45% = $140,625 saved

**Where the starting numbers come from:** Sites in our past ROI survey paid $125 to $250 per transfer. Use your own carrier or shuttle rate.

**Check it yourself:** Pull shuttle or transfer orders from the WMS or TMS. Flag pallets that went out and came back.

### Missed pickups

*Fees you stop paying.* When a load isn't ready, the truck leaves empty and you still pay.

**Formula:** Missed pickups × cost per miss × cut

**Proof:**

- asked for missed-pickup alerts by customer (Global confectionery manufacturer)

**The math:**

- Misses = loads × share missed or rebooked
- Cost = misses × cost per miss (TONU fee plus any extra for the new truck)
- Saved = cost × cut

**Illustrative example (the calculator's default inputs, not a customer result):**

- 12,500 loads × 2% = 250 misses
- 250 × $400 = $100,000 a year
- $100,000 × 40% = $40,000 saved

**Where the starting numbers come from:** ATS. Most TONU fees run $150 to $300 per truck.

**Check it yourself:** Count TONU charges on carrier bills. Add loads that shipped on a different carrier or a later date than booked.

### Yard moves

*Hours you don't work.* Every extra trailer shuffle is paid yard driver time.

**Formula:** Moves you could avoid × minutes per move × pay

**Proof:**

- 0.5–1 yard drivers saved per plant (Our model)

**The math:**

- Hours saved = yard moves × share you could avoid × minutes per move ÷ 60
- Saved = hours saved × floor pay

**Illustrative example (the calculator's default inputs, not a customer result):**

- 40,000 moves × 15% × 12 min ÷ 60 = 1,200 hours
- 1,200 hours × $28 = $33,600 saved

**Where the starting numbers come from:** Basic math. Pull moves and times from your yard system.

**Check it yourself:** Use the yard move log for each trailer. Count trailers moved more than once before unload or departure.

### Trailer fill

*Fees you stop paying.* Fuller trailers mean fewer loads for the same cases. Our apps flag underfilled trailers and show each lane how to fill more: consolidate loads, change trailer types, or change load and pallet types.

**Formula:** Loads avoided × cost per load

**Proof:**

- $2M+ saved across a complex network (F&B network)

**The math:**

- Loads after = loads × fill today ÷ fill after
- Loads avoided = loads − loads after
- Saved = loads avoided × cost per load

**Illustrative example (the calculator's default inputs, not a customer result):**

- 12,500 loads × 85% ÷ 86% = 12,355 loads
- 12,500 − 12,355 = 145 fewer loads
- 145 × $1,200 = $174,419 saved

**Where the starting numbers come from:** ATRI via Transport Topics. Trucks cost $2.336 a mile to run in 2025.

**Check it yourself:** Get space or weight use per load from your TMS or load-building tool. Sort lanes by average fill to find where to start.


## Travel

### Door choice

*Hours you don't work.* The wrong door makes forklifts drive farther for every pallet.

**Formula:** Extra driving hours × pay + extra miles × forklift cost

**Proof:**

- $29K driving saved a year, with $30k more to find (Top 5 global F&B manufacturer)
- 69 hrs extra driving in one day (Furniture retailer)
- 3,298 mi driving cut in 90 days (Global snack food manufacturer)

**The math:**

- Extra feet = pallets × share at the wrong door ÷ pallets per trip × extra feet per trip
- Driving hours = extra feet ÷ (forklift speed × 5,280 feet per mile)
- Driving cost = driving hours × floor pay + extra feet ÷ 5,280 × forklift cost per mile
- Saved = driving cost × share you win back

**Illustrative example (the calculator's default inputs, not a customer result):**

- 700,000 pallets × 40% ÷ 1 per trip × 65 ft = 18,200,000 extra feet (3,447 miles)
- 18,200,000 feet ÷ 21,120 feet per hour (4 mph) = 862 hours of driving
- 862 hours × $28 + 3,447 miles × $1 = $27,576 a year
- $27,576 × 60% won back = $16,545 saved

**Where the starting numbers come from:** Speed. 4 mph is a safe average with turns, stops and lifts. One customer site measured 5.4 mph. Distances come from your map of doors and storage.

**Check it yourself:** Compare the planned door to the door used for each load. Most WMSs log both. Add pallets per load and the distance from each door to each zone.

### Empty forklift driving

*Hours you don't work.* Forklifts that drive back empty are paid to carry nothing.

**Formula:** Empty driving hours avoided × (pay + forklift cost per hour)

**Proof:**

- −32% empty driving (Customer sites)

**The math:**

- Driving hours = forklift drivers × hours each × share of time driving
- Empty hours = driving hours × share driven empty
- Hours saved = empty hours × cut
- Saved = hours saved × floor pay + hours saved × forklift speed × forklift cost per mile

**Illustrative example (the calculator's default inputs, not a customer result):**

- 40 drivers × 2,000 hours × 50% = 40,000 driving hours
- 40,000 × 45% = 18,000 hours driving empty
- 18,000 × 30% = 5,400 hours saved (21,600 miles)
- 5,400 hours × $28 + 21,600 miles × $1 = $172,800 saved

**Where the starting numbers come from:** The empty share depends on your site. When no tasks are mixed, almost every trip back is empty. That puts it near 50%.

**Check it yourself:** Use task start and end spots from the WMS. Empty trips are the gaps between where one task ends and the next one starts.

### Low pick density

*Hours you don't work.* More picks at each stop means fewer stops and less walking.

**Formula:** Stops saved × seconds per stop × pay

**Proof:**

- $407K picking labor saved (Freight provider case study, https://autoscheduler.ai/wp-content/uploads/2023/07/AutoScheduler-CaseStudy-Top-Tier-Freight-Provider-0422.pdf)
- +30% pick rate at key sites (CPG case study, https://autoscheduler.ai/wp-content/uploads/2025/01/ASI_Gen_CS-Final-11-7-24-2.pdf)
- −10% time per pick (Top global 3PL)

**The math:**

- Stops today = picks ÷ picks per stop today
- Stops after = picks ÷ picks per stop after
- Hours saved = (stops today − stops after) × seconds per stop ÷ 3,600
- Saved = hours saved × floor pay

**Illustrative example (the calculator's default inputs, not a customer result):**

- 5,000,000 ÷ 1.3 = 3,846,154 stops today
- 5,000,000 ÷ 1.45 = 3,448,276 stops after
- 397,878 fewer stops × 45 seconds ÷ 3,600 = 4,973 hours
- 4,973 hours × $28 = $139,257 saved

**Where the starting numbers come from:** Dematic. Travel is more than half of picking time. Picking is about 55% of warehouse costs.

**Check it yourself:** Use pick records with location and time. Picks per stop = picks ÷ the number of different spots visited per trip.

### Missed cross docks

*Hours you don't work.* Pallets that go straight from receiving to shipping skip two moves.

**Formula:** Pallets cross-docked × handling time × pay

**Proof:**

- >5% of volume cross-docked (Consumer foods case study, https://autoscheduler.ai/wp-content/uploads/2023/08/AUT_CaseStudy_EnhancedAutoPilot-LowRes.pdf)
- +46.3% loads shipped straight from the plant (P&G)

**The math:**

- Minutes per pallet = 60 ÷ putaways per hour + 60 ÷ pulls per hour
- Hours saved = extra pallets cross-docked × minutes per pallet ÷ 60
- Saved = hours saved × floor pay

**Illustrative example (the calculator's default inputs, not a customer result):**

- 60 ÷ 30 + 60 ÷ 30 = 4 minutes per pallet
- 30,000 pallets × 4 minutes ÷ 60 = 2,000 hours
- 2,000 hours × $28 = $56,000 saved

**Where the starting numbers come from:** Sites in our past ROI survey did 25 to 40 putaways and 25 to 40 pulls per hour. Use your own.

**Check it yourself:** From WMS history, count pallets received and shipped the same day without a storage spot.


## Labor and output

### Units per hour (UPH)

*Hours you don't work.* When units per hour go up, the same work takes fewer hours. This one already includes travel, pick density and overtime.

**Formula:** (Hours today − hours after) × pay

**Proof:**

- 9–14% productivity per site (Food & bev plants case study, https://autoscheduler.ai/wp-content/uploads/2025/11/AUS-2025-Case-Study-AI-Driven-Flow.pdf)
- 9–12% productivity (PepsiCo)
- +18% productivity (Kellanova)

**The math:**

- Hours today = units ÷ UPH today
- Hours after = units ÷ (UPH today × (1 + UPH gain))
- Full-time workers saved = (hours today − hours after) ÷ paid hours per worker
- Saved = hours saved × floor pay

**Illustrative example (the calculator's default inputs, not a customer result):**

- 20,000,000 ÷ 120 UPH = 166,667 hours today
- 20,000,000 ÷ 132 UPH = 151,515 hours after
- 15,152 hours ÷ 2,080 = 7.3 full-time workers
- 15,152 hours × $28 = $424,242 saved

**Where the starting numbers come from:** Basic math. One trap: a UPH gain of g% saves g ÷ (1 + g) of hours.

**Check it yourself:** Use units and paid hours for each job from the WMS and time clock. Count them the same way before and after.

### Overtime

*Hours you don't work.* Overtime costs 1.5 times regular pay. Planning ahead moves it back to regular time.

**Formula:** Overtime hours avoided × overtime premium

**Proof:**

- 24–48 hrs planned ahead, less overtime (Food & bev network case study, https://9458072.fs1.hubspotusercontent-na1.net/hubfs/9458072/Case%20Study%20Food%20and%20Beverage.%20Network.pdf)
- −50% overtime (PepsiCo, Kellanova)

**The math:**

- Overtime hours = workers × paid hours each × overtime share
- Overtime avoided = overtime hours × cut
- Low case (moved to regular time) = overtime avoided × (overtime pay − floor pay)
- High case (the work goes away) = overtime avoided × overtime pay

**Illustrative example (the calculator's default inputs, not a customer result):**

- 150 workers × 2,080 hours × 8% = 24,960 overtime hours
- 24,960 × 50% = 12,480 overtime hours avoided
- 12,480 × ($42 − $28) = $174,720 if moved to regular time
- 12,480 × $42 = $524,160 if the work goes away

**Where the starting numbers come from:** US Department of Labor. Overtime is at least 1.5 times regular pay after 40 hours a week.

**Check it yourself:** Export regular and overtime hours by week and department from your time clock.

### Temp workers

*Hours you don't work.* Temps cost their pay plus the agency's markup.

**Formula:** Temp hours avoided × agency bill rate

**Proof:**

- sites staffed for peak plus absences (Our labor check)

**The math:**

- Temp hours avoided = temp hours × cut
- Saved = temp hours avoided × agency bill rate

**Illustrative example (the calculator's default inputs, not a customer result):**

- 60,000 hours × 10% = 6,000 temp hours avoided
- 6,000 hours × $27 = $162,000 saved

**Where the starting numbers come from:** Instawork. Warehouse agency markups usually run 25% to 40%.

**Check it yourself:** Use agency bills. Compare hours and bill rate by week to the volume those weeks.

### Automation output

*Room to grow.* Machines fed work in the right order get more done.

**Formula:** Extra units × cost to move them another way

**Proof:**

- +30–35% flow at automated sites (CPG case study, https://autoscheduler.ai/wp-content/uploads/2025/01/ASI_Gen_CS-Final-11-7-24-2.pdf)
- +35% product flow (Food & bev plants case study, https://autoscheduler.ai/wp-content/uploads/2025/11/AUS-2025-Case-Study-AI-Driven-Flow.pdf)
- +30% output in crunch periods (P&G)

**The math:**

- Extra units a day = machine hours per day × UPH today × output gain
- Extra units a year = extra units a day × days open
- Value = extra units a year × cost to move a unit another way

**Illustrative example (the calculator's default inputs, not a customer result):**

- 20 hours × 185 UPH × 13% = 481 more units a day
- × 300 days = 144,300 more units a year
- 144,300 × $1 = $144,300 a year

**Where the starting numbers come from:** The value of each unit depends on what you'd do instead. That could be overtime, a weekend shift or paying a 3PL to take the overflow.

**Check it yourself:** Use machine output by hour from your WCS. Look for gaps where a machine waited for work.

### Production line uptime

*Room to grow.* Plant lines stop when the warehouse can't keep up.

**Formula:** Line hours lost × cost per hour × cut

**Proof:**

- +35% product flow in plant warehouses (Food & bev plants case study, https://autoscheduler.ai/wp-content/uploads/2025/11/AUS-2025-Case-Study-AI-Driven-Flow.pdf)

**The math:**

- Hours lost = lines × minutes per day stopped by the warehouse ÷ 60 × days open
- Saved = hours lost × cost per line hour × cut

**Illustrative example (the calculator's default inputs, not a customer result):**

- 4 lines × 10 min ÷ 60 × 300 days = 200 line hours lost
- 200 × $2,000 × 40% = $160,000 saved

**Where the starting numbers come from:** Cost per line hour depends on your site. Finance usually has it as margin per hour of output.

**Check it yourself:** Use line downtime logs from your MES. Filter to stops caused by missing materials or full output space.


## Office time

### Wave building

*Hours you don't work.* Someone types waves in by hand every day.

**Formula:** Hours building waves × share automated × office pay

**Proof:**

- 10 min waves rebuilt, none by hand (E-commerce fulfillment 3PL)

**The math:**

- Hours a year = hours per day building waves × days open
- Hours automated = hours a year × share automated
- Saved = hours automated × office pay

**Illustrative example (the calculator's default inputs, not a customer result):**

- 4 hours × 300 days = 1,200 hours a year
- 1,200 × 90% = 1,080 hours automated
- 1,080 hours × $35 = $37,800 saved

**Where the starting numbers come from:** Time it yourself. Track one week of the person who builds waves, from start to release.

**Check it yourself:** Use wave build and release times by user from the WMS, or time it for one week.

### Cut research

*Hours you don't work.* Someone digs through the WMS to explain every short line.

**Formula:** Short lines × minutes each × office pay

**Proof:**

- −50% controllable cuts (CPG case study, https://autoscheduler.ai/wp-content/uploads/2025/01/ASI_Gen_CS-Final-11-7-24-2.pdf)
- site-built board shows what can cover each short (Top 5 global F&B manufacturer)

**The math:**

- Hours a day = short lines per day × minutes each ÷ 60
- Hours automated = hours a day × days open × share automated
- Saved = hours automated × office pay

**Illustrative example (the calculator's default inputs, not a customer result):**

- 60 lines × 6 min ÷ 60 = 6 hours a day
- 6 × 300 days × 75% = 1,350 hours automated
- 1,350 hours × $35 = $47,250 saved

**Where the starting numbers come from:** Time it yourself. Count a week of short lines and how long each one takes to fix.

**Check it yourself:** Pull short and cut lines by day from the WMS. Time one week to get minutes per line.

### Shift planning

*Hours you don't work.* A supervisor builds the crew plan before every shift.

**Formula:** Planning minutes × share automated × office pay

**Proof:**

- 80 hrs freed up each week (Consumer foods case study, https://autoscheduler.ai/wp-content/uploads/2023/08/AUT_CaseStudy_EnhancedAutoPilot-LowRes.pdf)
- 96% planning time saved (P&G)

**The math:**

- Hours a year = shifts per day × minutes planning each ÷ 60 × days open
- Hours automated = hours a year × share automated
- Saved = hours automated × office pay

**Illustrative example (the calculator's default inputs, not a customer result):**

- 3 shifts × 60 min ÷ 60 × 300 days = 900 hours
- 900 × 90% = 810 hours automated
- 810 hours × $35 = $28,350 saved

**Where the starting numbers come from:** Time it yourself. Track the minutes from "what's coming today" to a posted crew plan.

**Check it yourself:** Time each shift's planning for one week.

### Spreadsheets and reports

*Hours you don't work.* People rebuild the same trackers and reports every day.

**Formula:** Tracker hours × share automated × office pay

**Proof:**

- $100K+ a year from one site-built app (Top 5 global F&B manufacturer)
- 11 min from report to live app (Top 5 global F&B manufacturer)

**The math:**

- Hours a year = people × hours per week each × weeks
- Hours automated = hours a year × share automated
- Saved = hours automated × office pay

**Illustrative example (the calculator's default inputs, not a customer result):**

- 6 people × 4 hours × 50 weeks = 1,200 hours
- 1,200 × 80% = 960 hours automated
- 960 hours × $35 = $33,600 saved

**Where the starting numbers come from:** Time it yourself. Ask each coordinator which files they open every shift and how long they spend in them.

**Check it yourself:** List the shift's spreadsheets and reports, who keeps each one and the hours per week.
