almohannad.cx
Marketing Research & Metrics

American Customer Satisfaction Index

6 min read Translated from the Arabic original

Introduction

I mentioned the American Customer Satisfaction Index (ACSI) in a previous article as a credible reference that can be relied upon when you want to use benchmarking indicators — provided its use is widespread enough in your region to enable meaningful local comparisons.

The ACSI is a scientific model developed by academic specialists at the University of Michigan in the United States. It first saw the light of day in 1994, and to date it has spread to roughly sixteen countries — in the Middle East, only Kuwait and Saudi Arabia. In this article I’ll cover the model, its components, and the recent expansion of its operations to include the Kingdom of Saudi Arabia.

Model Details

The model is grounded in cause-and-effect principles and consists of several mathematical equations falling under econometrics and statistics. The detailed equations behind models of this type are, in most cases, not publicly disclosed. Like many economic or marketing-research models, it functions as a “black box” whose internals are revealed only in very rare cases and under strict criteria, to bodies that evaluate them. Yet the reputation of models with this profile becomes strong through user testimonials and repeat usage. More detail on the mathematical equations of this model is available in an older, copyrighted article I can’t quote from without formal permission — you can refer to it below:

The American Customer Satisfaction Index: Nature, Purpose, and Findings

Model Components

Perceived Quality

The degree to which the product or service meets the customer’s needs, the customer’s confidence in it, and the frequency of problems experienced with that product or service.

Customer Expectations

The customer’s impressions of the company’s products or services being researched. These are expressed through the customer’s prior experiences with the company, or through non-experiential impressions such as a friend’s recommendation or impressions left by advertising. The customer is also asked about their expectations regarding the company’s future ability to deliver the hoped-for quality.

Perceived Value

This component examines the impact of the price paid by the customer, and whether they feel the quality they received was worth what they paid. They’re also asked about the likelihood of doing business again in the future.

Customer Satisfaction Index

This index represents the weighted average satisfaction of the three components above. The model’s operators use specialized software to determine the importance of each component, producing an importance weight for every aspect. This weight is then applied to the satisfaction levels to derive weighted satisfaction.

Customer Complaints

Complaints have a negative relationship with satisfaction. They are measured as a percentage of customers who lodged a complaint directly with the company within a defined timeframe.

Customer Loyalty

A leading indicator of profitability. Loyalty in this model is measured by the likelihood of the customer repurchasing from the same provider in the future, and their willingness to purchase that provider’s products or services at different price points.

The model components that, when processed, lead to the satisfaction index can vary from industry to industry — I’m referring here to perceived quality, perceived value, and customer expectations. For example, the following components are part of a model designed specifically for banks: ATM locations, branch locations, customer service representatives, fees and tariffs, the website, and certain premium products.

The survey used to collect the data the model requires contains a set of standardized questions that cannot be changed — though you can add to them. The survey is designed to cover the most important aspects of the customer experience and differs from one industry to another depending on how the customer experience differs across them. The scale used in the survey is a ten-point scale, later transformed during analysis to be displayed on a 100-point basis.

This is why this type of study can be considered syndicated research when all competitors in a given industry participate. Is that even possible? Well, it can be voluntary — but in most cases, what drives participation in this type of research is government bodies or professional unions that oversee the performance of specific industries and impose performance-measurement procedures to create a healthy competitive environment.

Studies using this model are typically run on a quarterly basis, though this varies by industry — some industries change so rapidly that the frequency of studies must increase. The research naturally targets the current customers of the entity being studied. In some cases, customer lists are obtained directly from the organizations sponsoring the study; in others, customers are recruited via a screening survey that precedes the main one. As for sample size, the model’s experts state that it requires no fewer than 250 targeted respondents per study to produce reliable results. I once read a case study about a project they ran in the United States and they made do with that same sample size distributed across 48 different states. Statisticians and marketing researchers may find this sample size surprising, given the implied margin of error — but I can’t pass judgment until I review the weighting methodology, which I haven’t yet had a chance to do.

Results and Reporting

The main outputs from research using this model are presented in a report — much like any other marketing research. The insights, conclusions, and decisions drawn from reading the report will vary by reader. Generally, the most important points are:

  • Performance comparison against competitors: provided they ran the same study using the same model in the same timeframe. Regional or global comparisons can also be made within the same industry, as well as domestic comparisons — not necessarily against competitors but against leading companies.
  • Areas for improvement: aspects that scored low on satisfaction while also having a high impact on satisfaction.
  • Satisfaction level produced by the model: the model’s operators have established a direct relationship between the result and the company’s financial performance, as well as its stock price.
  • Growth forecasts: the model also provides indicators that predict expected growth — critical for decision-makers and strategy-setters.

Entry into the Kingdom of Saudi Arabia

On March 12 of this year, an introductory seminar was held at the Hyatt Park hotel in Jeddah announcing the availability of this model’s applications here in Saudi Arabia. The seminar was organized by Khotwa & Tahleel (Plan & Analysis), owned by Mostafa Naqwi, a veteran in marketing research and one of the longest-serving practitioners in this field in Saudi Arabia. The company’s head office is in Jeddah, and they have secured an exclusive contract with the model’s operators in the United States to run it in Saudi Arabia — so any client wishing to use this model must work through them. While the field execution of the research — interviews and data collection — will be carried out by the company, in most cases the data analysis and results extraction for this type of model take place in the model owners’ home country (the United States), inside the black box I mentioned earlier. I’m not certain this also applies to this particular model, but it’s what I’d expect.

What about the model’s drawbacks and weaknesses?

I’ll wait for a hands-on personal experience with this model in the future to be able to critique it from lived experience. As for academic or scientific critique, I’m not qualified to offer it.

You can learn more by visiting the model owners’ official site or the blog dedicated to it.

In closing: I welcome any reader’s critique. Constructive criticism ripens ideas.

Related framework notes: NPS, CSAT, CES, Voice of Customer (VoC)

← Articles MOC