The American Customer Satisfaction Index (ACSI): Quarter 3, 2025
A Threat Potentially More Damaging Than the Great Recession:
The Decoupling of Seller Profits from Buyer Utility
In addition to the government shutdown, economic data blackouts, and volatile tariff rate changes, there is another, yet mostly unnoticed, threat to the very functioning of U.S. economy: High profits, record-breaking stock returns, and weakening customer satisfaction. The American Customer Satisfaction Index (ACSI®) Q3 results are a warning sign. While the national score is unchanged at 76.9 (out of 100), it follows a steep decline and long-term stagnation that is a slow-moving threat which, if not reversed, might inflict severe damage to economic growth: the decoupling of buyer utility from seller profit.
Consumer switching costs have risen over the past decade, especially in digital and service sectors, with less competition as a result. Online retail, video streaming, and search engines have decreasing customer satisfaction this quarter. Overall, markets have become more concentrated with rising seller pricing power. Corporate profits as a share of national income have increased as well, so have customer complaints. Mergers and acquisitions have escalated, while antitrust enforcement has not. Consumer surplus is eroding, with less household purchasing power as a result. If economic history is a guide, innovation will suffer, capital allocation will be inefficient, and productivity growth will slow. There will be rising inequality and further redistribution of income from consumers to shareholders. Because high-income groups have lower marginal propensity to consume, consumer demand will fall and both groups will be worse off.
The stock returns of the ACSI companies with the highest scores in their respective markets point in the same direction. Not long ago, these companies produced significantly higher returns relative to the S&P 500. Over the long term, from 2006 to the end of September 2025, that is still true, but recently there has been a dramatic change. The asset-weighted S&P 500 has become extremely top-heavy, with fewer than 2% of the companies sometimes accounting for more than 50% of its return. The equal-weight S&P counterpart, which used to outperform the asset-weighted one, now has returns slightly below the top ACSI companies, while both have returns below the S&P 500.
There is no question as to the remedy: Sellers should compete for buyers. While the notion of the perfect competitive market equilibrium where buyers maximize utility and sellers maximize profit is an idealistic unobtainable reality, its opposite is a tangible threat: When sellers make profit because buyers have limited choice, the economy suffers. Sellers are supposed to strengthen buyer relationships by providing buyer satisfaction superior to that of their competition – if they do, they would be rewarded. If they don’t, consumer and equity markets render punishment. That is how a market economy is supposed to work, but that is not what we have today. There are exceptions, however, and they may be worth mentioning: Apple, Google, Microsoft, and Costco are longtime front-runners in customer satisfaction. Most of them belong to the high stock-return technology sector and have pricing power, but they have not exercised that power to the detriment of their customers. Accordingly, even their short-term three- and five-year stock returns are higher than the S&P 500 return.
A healthy market economy relies on constructive competition among sellers for the satisfaction of buyers, with prices aligned with costs and demand rather than monopoly power. Increasing customer satisfaction implies that consumers realize value, with rising surplus and high marginal utility of expenditure as a result. Most data suggests that we now have the opposite. The reason that it might be more damaging than the Great Recession of 2007-2009 is that it threatens the very functioning of a market economy, which was not the case back then.
Claes Fornell is the Donald C. Cook Distinguished Professor of Business Administration (Emeritus) at the Ross School of Business, University of Michigan.
| 1st Quarter | 2nd Quarter | 3rd Quarter | 4th Quarter | |
|---|---|---|---|---|
| 2025 | 77.0 | 76.9 | 76.9 | |
| 2024 | 78.0 | 77.9 | 77.9 | 77.3 |
| 2023 | 75.4 | 76.7 | 77.1 | 77.8 |
| 2022 | 73.1 | 73.0 | 73.5 | 74.4 |
| 2021 | 73.9 | 73.8 | 73.5 | 73.1 |
| 2020 | 74.3 | 74.1 | 73.9 | 73.6 |
| 2019 | 75.6 | 75.7 | 75.7 | 75.2 |
| 2018 | 76.7 | 76.2 | 75.9 | 75.6 |
| 2017 | 76.6 | 76.9 | 76.9 | 76.9 |
| 2016 | 76.3 | 76.2 | 76.4 | 76.7 |
| 2015 | 76.2 | 76.1 | 76.1 | 76.1 |
| 2014 | 76.5 | 76.4 | 76.5 | 76.5 |
| 2013 | 76.6 | 76.5 | 76.7 | 76.8 |
| 2012 | 75.9 | 75.9 | 75.9 | 76.3 |
| 2011 | 75.6 | 75.7 | 75.7 | 75.8 |
| 2010 | 75.9 | 75.9 | 75.7 | 75.3 |
| 2009 | 76.0 | 76.1 | 76.0 | 75.9 |
| 2008 | 75.2 | 75.1 | 75.0 | 75.7 |
| 2007 | 75.2 | 75.3 | 75.2 | 74.9 |
| 2006 | 74.1 | 74.4 | 74.4 | 74.9 |
| 2005 | 73.0 | 73.1 | 73.2 | 73.5 |
| 2004 | 74.4 | 74.4 | 74.3 | 73.6 |
| 2003 | 73.8 | 73.8 | 73.8 | 74.0 |
| 2002 | 73.0 | 73.0 | 73.1 | 72.9 |
| 2001 | 72.2 | 72.1 | 72.0 | 72.6 |
| 2000 | 72.5 | 72.8 | 72.9 | 72.6 |
| 1999 | 72.1 | 72.0 | 72.1 | 72.8 |
| 1998 | 71.9 | 72.2 | 72.3 | 72.6 |
| 1997 | 70.7 | 71.1 | 71.1 | 70.8 |
| 1996 | 73.0 | 72.4 | 72.2 | 72.0 |
| 1995 | 74.1 | 73.7 | 73.7 | 73.7 |
| 1994 | – | – | 74.8* | 74.2 |
*Baseline measurement taken in summer 1994
While companies today have more data about their customers, the analytics employed to turn data into information are for the most part not good enough. Customer satisfaction data have certain characteristics that make it difficult to obtain accurate estimates, to pinpoint what aspects of the customer experience need attention, and to gauge the financial impact of actions contemplated. Traditional statistical methods assume normal frequency distributions among the residuals, moderate multicollinearity, and low levels of data noise. Customer satisfaction data don’t meet these assumptions.
ACSI Analytics is designed to overcome these problems and thereby turning raw data into financially relevant information by:
- Separating signals from noise
- Moving from correlations and artificial intelligence (AI) patterns to cause-and-effect interpretations
- Calibrating measurement instruments toward profitability
Data is not the same as information—especially not data from consumer surveys. Management decisions require information; raw data must be filtered in order to be useful for decision-making. ACSI technology filters out data noise.
Management decisions require cause-and-effect information—something that current CX tools, whether based on AI or descriptive statistics, don’t provide. ACSI Analytics, on the other hand, is based on a causal model.
There is a wide disparity in the amount of consumer data collected by companies today. Some data suppliers use surveys with more than 200 questions per respondent, while others focus on responses to a single question. Neither is appropriate. Excessively long surveys may lead to straight-line responses. Good measurement techniques—whether in the social or physical sciences—typically require several measures (survey questions in this case) per product feature or service dimension.
Accuracy and relevance are what matters. To contribute to the business objectives at hand, the measurement instruments need calibration in ways similar to the physical sciences. This is why companies with high scores in the American Customer Satisfaction Index, which is calibrated to maximize customer loyalty, are financially successful, most notably in terms of stock returns and profitability.
No advertising or other promotional use can be made of ACSI data and information without the express prior written consent of ACSI LLC.