Delivery Window Width: How Wide Should a Promise Be? A 2026 Model of Cost per Stop

Delivery window width is the length of the time period a customer is given for an attended delivery, such as one hour, two hours, four hours or the whole day. It is a design variable with a cost curve and not a customer-service setting, because every hour taken out of a window removes options from the route planner. The research and the simulation below agree on the shape of the curve: two-hour and three-hour windows cost modestly more than four-hour windows, one-hour windows cost far more, and the right answer is usually a four-hour default with narrower windows sold as a priced option.

Why Delivery Window Width Matters: The Business Case

Narrow windows are popular with customers and expensive for routes, so the evidence has to be read on both sides.

• Agatz, Campbell, Fleischmann and Savelsbergh wrote in Transportation Science that delivery cost rose by up to 25% when two-hour slots replaced slots spanning an entire morning or afternoon, using data from a Dutch e-grocer.
• The same paper recounts earlier research finding that completely flexible, unattended delivery reduced costs by up to a third compared with attended delivery in two-hour slots.
• Köhler, Ehmke and Campbell’s 2020 paper in Omega studies flexible time window management because short windows can increase delivery costs significantly.
• An MIT Sloan Management Review study of 152,195 online grocery transactions at a European retailer found customers would wait 10.8 hours longer for a delivery window one hour shorter, which shows why customers value narrow windows.
• Locus’s own test of Delivery Promise Management against a retailer’s tightened-capacity order book found that 98.5% of orders received a window the network could keep.

Benchmarks for failed first attempts vary so widely by source and definition that no single figure is safe to quote, so this guide uses assumptions you can replace with your own rate.

How to Choose a Delivery Window Width in Six Steps

The six steps put a price on every hour of window width.

1. Measure what each window width costs on your network

Re-plan a sample of real days with each window width, holding stops, vehicles and service times fixed. The cost per stop for each width is your own cost curve.

2. Measure failed attempts by window width

Track failed first attempts by the window the customer was given. Wider windows usually raise failed attempts because customers cannot stay in for as long.

3. Add the failed-attempt cost

Multiply each width’s failed-attempt rate by the cost of a redelivery and add it to the route cost per stop. The total is the cost of promising that width.

4. Find where the curve bends

Compare adjacent widths. In most networks the cost steps are small between three and six hours and steep below two hours, and the bend tells you where the default should sit.

5. Price the narrower windows

Offer the default width at no charge and sell narrower windows at a price at or above their extra cost per stop, with a limited number of slots per route.

6. Re-test when density or fleet changes

Dense territories absorb narrow windows more cheaply than sparse ones, so repeat the test by territory and when volumes change.

One-Hour, Two-Hour, Four-Hour and All-Day Windows: Key Differences

What to Look for in Delivery Slot Software

Capacity-checked slots

The software should check every slot against live route capacity before the customer sees it, so a window shown at checkout is a window the network can serve.

Slots by service level

Look for feasible windows returned by service level, so you can price the premium ones and keep the default free.

Re-planning that protects booked windows

A new plan should go live only when every promised window still holds, so changes during the day do not break commitments.

Reasons for refusals

When a slot is refused, the system should say why, such as the route being full or the cut-off having passed, so planners know what to fix.

Window width as a configurable constraint

You should be able to set window width by territory, service level and day, and test the effect in a simulation before changing what customers see.

Delivery Window Width in Action: Real-World Results

A simulation on synthetic data. The setup is an assumption set and not a benchmark. A depot sits in the center of a 25 km by 25 km area with 120 stops, 35 km/h average speed, 6 minutes of service per stop, a 40-stop vehicle limit and a 10-hour route limit. Windows start on the hour between 09:00 and 19:00. A greedy insertion heuristic builds routes. Cost is $90 per vehicle-day plus $0.70 per km plus $28 per paid hour, and the table averages 100 random instances.

The simulation’s gap between two-hour windows and morning or afternoon windows, roughly 15% to 40% depending on the comparison, brackets the up-to-25% figure Agatz and colleagues reported, which gives some comfort that the curve has the right shape. The all-day result is cheapest on routes, but wide windows raise failed attempts. With assumed failed-first-attempt rates of 4%, 6%, 8%, 10%, 14% and 22% for the six widths and an $11 redelivery, the total cost per stop becomes $37.26, $22.58, $20.78, $20.17, $17.12 and $13.01.

The steps between adjacent widths are the pricing signal. Moving from four hours to two hours costs about $2.41 more per stop including failed attempts, and moving from two hours to one hour costs about $14.68 more. A one-hour window therefore needs a price near $15 or a very dense route, while a two-hour option can be sold for a few dollars.

What a minority of narrow windows costs. The same instances were re-run with a mix of widths. Giving 10% of customers a one-hour window and the rest four hours changed route cost by +$0.03 a stop, with a standard error of $0.20. Giving 20% two-hour windows changed it by +$0.73, and giving 50% two-hour windows by -$0.46, which is within the noise of a greedy heuristic. Making two-hour windows the default for everyone added $2.85. The lesson is that narrow windows are cheap as an option and expensive as a default. At double the stop density in the same area, the two-hour premium over four hours fell to about $1.50.

A Vietnamese beverage distributor. The distributor runs a single depot out to thousands of small retail points a day on a mixed fleet, where every shop has its own opening hours and access constraints. Manual planning took an hour to an hour and a half before any vehicle moved. After moving to Locus, it cut fuel consumption by 37%, raised orders per delivery trip by 22%, cut route distance by 12% and cut route planning time by 35%. Read the beverage distributor case study.

A Canadian grocery brand. The brand delivers fresh, perishable food to homes in more than 30 cities through several third-party carriers, where every order races a freshness clock. After Locus automated carrier selection, shipment creation and tracking through proof of delivery, it reported a 33% faster order-ready-to-delivered cycle time, a 15% reduction in fulfillment costs and 25% less time on manual shipping tasks. Read the Canadian grocery case study.

Common Delivery Window Width Mistakes to Avoid

Making the narrowest window the default. A two-hour or one-hour default prices every customer at the premium rate, so offer it as an option.

Using one average cost for all territories. A dense urban zone and a sparse suburb have very different cost curves, so test widths by territory.

Ignoring failed attempts when widening windows. Wider windows lower route cost and raise failed attempts, so count both before moving to a wider default.

Promising windows the network cannot serve. A window shown at checkout without a capacity check turns into a missed promise, which is why slots should be checked against live capacity first.

How Locus Approaches Delivery Window Width

Locus, the world’s first Decision-Intelligent, Agentic TMS, treats the delivery window as a decision made against live capacity and not a static setting. Its Delivery Promise Management checks every slot against the capacity behind it, from the origin cut-off to the vehicle at the door, and returns feasible slots in under a second by service level so premium windows can be priced.

Recognized by analysts and selected after evaluation. Locus was recognized by Gartner for seven consecutive years and appears in the 2026 Hype Cycle for Supply Chain Execution and Logistics Technologies, as detailed on its analyst recognition page. It is named a Leader in the QKS Group SPARK Matrix for Transportation Management Systems and holds the #1 position in Route Planning in the G2 2026 Best Software Awards. In October 2025, Ingka Investments, the investment arm of Ingka Group, the world’s largest IKEA retailer, acquired Locus. Ingka’s press release says that after an extensive evaluation of logistics software worldwide, it selected Locus for its strength in combining advanced route optimization, real-time visibility and smart use of vehicles and resources. Locus continues to operate independently.

What this means for window width. Locus plans routes against 250+ real-world constraints, including time slots, through its Route Planning System, and Delivery Promise Management tests every re-plan so a new plan goes live only when every promised window still holds. In Locus’s test against a retailer’s tightened-capacity order book, orders left without a slot fell from 3.6% to 1.5% once live re-optimization was running. That is a Locus test and not a customer outcome. Locus has optimized 1.5B+ deliveries for 360+ enterprise customers across 30+ countries.

What to confirm. Locus’s public pages describe slot checking and pricing by service level but do not publish a recommended window width or a cost curve. Ask for a simulation on your own orders with two-hour, three-hour and four-hour windows, and compare it to the curve in this guide.

To model your own window widths against live capacity, schedule a demo.

For retailers asking how wide a delivery window should be, Locus offers capacity-checked slots, service-level pricing and constraint-based planning in one platform, so window width is chosen from a cost curve and not a guess.

Frequently Asked Questions

What is delivery window width? It is the length of the time period a customer is given for an attended delivery, such as one, two or four hours. It decides how much freedom the route planner has to sequence stops.

How much do narrower delivery windows cost? Research by Agatz and colleagues found up to 25% higher delivery cost for two-hour slots than for morning or afternoon slots at a Dutch e-grocer. In the simulation here, two-hour windows cost about $2.85 more per stop than four-hour windows and one-hour windows about $17.76 more.

What is the best default window for attended deliveries? For most networks it is about four hours, with two-hour windows sold as an option. Test your own curve, because dense territories can support narrower defaults.

Should a retailer charge for narrow windows? Yes where the extra cost per stop is material. In the example, a two-hour window costs about $2.41 more per stop than a four-hour window including failed attempts, and a one-hour window about $14.68 more than a two-hour window.

Do wider windows cause more failed deliveries? Usually, because customers cannot stay in for as long. Published first-attempt benchmarks conflict by source and definition, so measure your own failed-attempt rate by window width.

Does Locus support delivery windows by service level? Locus’s Delivery Promise Management returns feasible slots by service level in under a second and checks each against live capacity. Its public pages do not publish a recommended window width.