Increase Delivery Success with Dynamic ETA Windows

When customers buy online, they expect full visibility into their delivery progress. 

 Yet many organisations struggle to confidently provide an estimated time of arrival (ETA). This leads to costly failed deliveries and customer dissatisfaction. 

 So how can you empower customers with accurate, precise ETAs? 

In this article, we'll ask what kind of information is most useful for customers. Then we’ll look at how dynamic ETAs improve the customer experience and increase delivery success.   

In this article: 

• What makes a good delivery ETA? 
• Why are ETA predictions a challenge? 
• How to improve delivery ETA predictions 
• Communicating delivery ETAs to customers 

What makes a good delivery ETA? 

There are two ways to measure the quality of an ETA: precision and accuracy.  

• A precise  ETA is one that’s narrowly defined (“2:00 pm - 2:15 pm”) or exact (“2:13 pm”). 
• An accurate ETA is one that's correct. The delivery arrives when you promised it would.   

Both of these factors are important, but not in the way you might think. 

To start with, the demand for precision varies based on the product category. Descartes' latest research shows that for smaller purchases, consumers are more concerned with cost than with the precision of the delivery window. 

For large purchases such as furniture and white goods, precision is more important. The same goes for B2B deliveries such as construction supplies: goods must arrive by a certain time, and staff must be able to attend the delivery. 

You’d think that most customers want a precise ETA window well before the day of delivery. However, what they really need is enough information to plan ahead, followed by real-time updates as the delivery gets closer. 

Imagine the scenario:  

A retailer has notified you that your order is to be delivered next Tuesday. You note the date in your diary and don’t think about it much for the next few days. 

What’s useful to you in this scenario is if the business can provide an accurate arrival window (say, 11 a.m. - 3 p.m.) in advance so that you can plan your week. 

This can be followed by a narrow two-hour window on the morning of the delivery, and precise real-time updates once the driver is a few stops away. This ensures that you’re available to attend the delivery, but don’t have to spend the whole day waiting around. 

Why are ETA predictions a challenge? 

Businesses struggle with estimated arrival times because they rely on predictions. There’s simply no way to be 100 percent certain of when a delivery will arrive.  

In complex routes with multiple stops, there are many variables that will impact service and travel times. Traffic, access issues and driver non-compliance can all cause unexpected delays.  

In these circumstances, providing a precise, but inaccurate, ETA may do more harm than good. 

So how can you give customers the information they need, without making unrealistic delivery promises? 

How to improve delivery ETA predictions 

The trick with delivery ETAs is to provide real-time updates as and when more information becomes available. 

Instead of rushing to provide an hour-and-minute ETA too soon, ask yourself:  

"What is the most precise range that we can confidently share with the customer at this time?"  

Machine learning models help to answer this question. Using real-time and historical data, it's possible to predict ETAs with increasing precision as the driver moves along their route. 

You can set parameters based on what level of confidence you're comfortable with. On the morning of a delivery, you might be able to share a four-hour arrival window with 85% confidence, or a two-hour arrival window with 80% confidence. Depending on your customers' expectations, you can decide which option is most appropriate. 

As the route progresses you can narrow the ETA range until it's possible to provide a 15-minute window or a precise hour-and-minute estimate. Normally, this will be when the customer's stop is next. 

These ETAs can be sent via API to a customer engagement platform that automates SMS or email communications. Customers receive updates at useful moments in the lead-up to their delivery. 

Communicating delivery ETAs to customers 

Here's what real-time ETA updates can look like in practice. Keep in mind that the ideal timings will vary depending on your product type, customer preferences and driver behaviour. 

1. Delivery confirmation  

When the order is confirmed, send the customer an ETA window e.g. Tuesday from 11 a.m. to 3 p.m. By keeping the estimate broad at this stage, you avoid making delivery promises that you can't keep. 

Reassure the recipient that you will share a more precise estimate as soon as you can. This gives customers peace of mind and helps to prevent "Where is my order? (WISMO) calls.

2. Delivery reminders 

It’s a good idea to send a reminder the day before the delivery, re-iterating the scheduled timeslot or providing a narrower estimate if possible. This increases first-attempt deliveries by allowing customers to plan ahead. 

On the morning of the scheduled delivery, you may be able to provide a more precise one-hour time window. The message can include a link to an online portal displaying the delivery progress in real time.  

Let the recipient know that you’ll send another message when the driver is fifteen minutes away. Customers can go about their day without wondering if their delivery is about to arrive.  

3. Real-time tracking 

When the driver is on their way, provide the customer with a precise ETA based on the vehicle’s location and predicted travel time. 

You can choose the appropriate level of information to share. This could be a combination of: 

• Place in the queue (“The driver is 3 stops away”; “Your stop is next") 
• Narrow arrival window (“Delivery expected in 15 mins”) 
• Hour-and-minute ETA (“Delivery expected at 2:35 p.m.”) 
• Approximate driver location on a map 
• Driver location on a map, with a route line 

This ensures that customers are ready and waiting to attend their delivery. 

4. Arrival notification 

When the delivery is about to arrive, you can send another SMS to help the customer prepare. The SMS can be triggered based on a time-fence, when the driver is five minutes away. 

This is especially useful for bulky deliveries where the customer may need to clear access. 

Getting started with dynamic ETA windows 

Consumers demand full transparency around deliveries. But it’s important to get the balance right: Too little information leaves customers in the dark but providing a precise ETA too early can backfire.  

With dynamic ETAs, you can keep customers in the loop without making unrealistic delivery promises.  

Descartes’ machine learning tools determine the best time-range to share with the customer, based on real-time and historical data. This information can trigger customer notifications through your existing communication solution, or via the Descartes Customer Engagement™ platform.

The benefits go beyond customer experience. By proactively communicating with customers, your business stands to reduce inbound calls and increase first-time delivery rates. 

Contact our last-mile delivery experts to get started.