Apologizing on Demand : A Field Experiment on the effects of apology on future demand

How to cite

Ho, Benjamin. 2017. "Apologizing on Demand : A Field Experiment on the effects of apology on future demand." AEA RCT Registry. AEARCTR-0002342. https://doi.org/10.1257/rct.2342-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Benjamin Ho*

* Primary investigator

Registry ID
AEARCTR-0002342
DOI
https://doi.org/10.1257/rct.2342-2.0
First registered
Last updated
Status
In development
Trial dates
2017-07-20 – 2017-12-31
Countries
Private
JEL codes
D91, C93
Registry record
http://www.socialscienceregistry.org/trials/2342

Abstract

Apologies are a key customer service tool for managing customer relationships. We run a field experiment with a retail technology company where customers who have a bad experience are offered an email apology and monetary compensation. We vary the wording of the apology and the size of compensation in order to test predictions of the apology model from Ho (2012). In particular, we are looking for the impact of an apology on the shape of future demand. Specifically, we want to know how apologies affect customer response to future price increases and decreases and to future bad experiences.

Intervention

The Intervention consists of identifying customers who have had a bad experience. After the bad experience, customers in the treatment group will receive an email an apology whose wording reflects either a generic apology, a reputation based apology, or a promise based apology (Ho, 2012). Each treatment group will be offered either $0 or $5 as compensation.

Experimental design

Customers who have a bad experience are randomized into one of the three apology conditions (generic apology, reputational apology, promise apology), and one of the two compensation conditions plus a no apology control. Within each of the apology conditions, subjects are randomized between receive $0 in compensation and $5 in compensation.

Design details

Randomization method
Randomization by computer
Randomization unit
Randomized at the individual level
Clusters
1.6 million individuals
Observations
1.6 million
Arms
200,000 per treatment arm (3 types of apologies + 1 non apology X $0 or $5 compensation) balanced for • Product usage in the previous month • Lifetime product usage • Product usage composition • Tenure on platform • Frequency of bad experiences in the past month • Tickets opened • Average past price
Primary outcomes
Primary outcomes are estimates of future demand as a function of the apology treatment (e.g. responsiveness of demand to future price changes, future circumstances, and future bad experiences as well the level of future demand).

Cited by

No citation data available for this registration yet.

Keywords