Beth Goody
If you’ve ever watched the minutes tick by while waiting for a delivery—perhaps dinner after a long day, or a package that you urgently need—you know the frustration you feel when the provider’s estimated time of arrival comes and goes but the meal or product you ordered is nowhere in sight.
You might have felt differently if the seller had given a time window rather than a single ETA, suggests research by Chicago Booth’s Beidi Hu, University of California at Berkeley’s Celia Gaertig, and Booth’s Berkeley J. Dietvorst. The way arrival time is expressed, they find, can influence a customer’s satisfaction.
Some businesses give point-in-time estimates (“your food will arrive in 35 minutes”), while others use a range (“your food will arrive in 30–40 minutes”). Whichever method they use, the ETA carries uncertainty, since delivery status is subject to variables such as prep time, worker availability, traffic, inventory shortages, and flight delays.
The researchers ran a series of studies to test consumer preference for point-in-time versus range estimates under various conditions. In one set of studies, the app provided online participants with either a point-in-time estimate (such as 45 minutes) or a time range that was narrow (40–50 minutes), moderate (35–55 minutes), or wide (15–75 minutes) for their food delivery. In each group, the participants then viewed a series of 20 “actual” arrival times and rated the app on a number of variables, including how much they trusted its estimates.
When is my meal coming?
Study participants were generally more satisfied with a food delivery app when given a time span rather than a specific point for when to expect their meal. But they didn’t like a time span that was too wide to be meaningful.
Participants didn’t like wide time ranges, the findings make clear. “Widening a range presents a trade-off between accuracy and informativeness,” the researchers write, explaining that people “seem to only prefer ranges to point estimates as long as the accuracy gained from providing (or widening) a range does not lead to too much information being lost.”
But when ranges were tighter, participants nearly always preferred those over point-in-time estimates. In a set of follow-up studies, the researchers looked into why ranges were so much more desirable and find that the answer has to do with expectations.
Before knowing the actual arrival time, participants who were given time ranges anticipated a wider delivery window than those who were given point-in-time estimates. When told that a delivery had arrived after the originally estimated time, participants who had seen a moderate time range to begin with were less likely to rate the delivery as “unexpectedly late.” Range estimates, write the researchers, “make outcomes less likely to violate people’s expectations,” perhaps because they signal uncertainty that consumers build into their plans.
People consistently appreciated when inherent uncertainty was communicated. In another study, participants preferred accurate time ranges over point-in-time estimates, even when the latter represented the top of the range—for instance, a 55-minute estimate was liked less than a range of 35–55 minutes. The researchers conclude that “uncertain estimates in the form of ranges are not disliked—they are actually preferred.”
For restaurants or consumer goods companies that simply want to enhance customer trust, the findings indicate that presenting a time frame (that’s not too wide) rather than a time point is the way to go.
Beidi Hu, Celia Gaertig, and Berkeley J. Dietvorst, “How Should Time Estimates Be Structured to Increase Customer Satisfaction?” Management Science, forthcoming.
Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.