In 1985, Costco's then chief operating officer, Craig Jelinek, told co-founder Jim Sinegal that the company was losing money on its $1.50 hot-dog-and-soda combo special and asked to raise the price. Sinegal's reply was an operating rule, not a joke: "If you raise the price of that xxx hot dog, I will kill you. Go figure it out."
Jelinek figured it out. Costco brought production in-house, cut cost, and left the customer promise intact. The combo has stayed at $1.50 since 1985.
Over the years, the same sequence has been applied. a hard cap on merchandise markups, generally 14 to 15%; wages and benefits that keep turnover low and service high; a membership system that pays the firm only if people keep coming back. Sinegal's order of priority was explicit—obey the law, take care of customers, take care of employees, respect suppliers, and only then reward shareholders.
Costco is now among the world's largest retailers. Long-term shareholder returns sit well above the S&P 500. Profit followed the customer constraint. It was not the starting purpose.
The tech giant, IBM, under CEO Sam Palmisano chose the opposite path. The IBM 2010 and 2015 earnings roadmaps made doubling earnings per share on a timetable the firm's top priority. Customers were barely in the picture.
Over the next decade, IBM spent well over $100 billion buying its own shares—often more than capital expenditure and research and development combined. Buybacks shrank the share count, so earnings per share could rise while revenue did not. Executive pay was tied to those targets. Missing something for customers carried no penalty.
Revenue then declined for more than 20 consecutive quarters in the 2010s. Hardware businesses were sold. Debt helped the scoreboard look healthy. Amazon, Microsoft, and Google took work IBM used to own. The $34 billion purchase of Red Hat arrived after customers had already gone elsewhere. Today, IBM's long-term returns remain well below the S&P 500.
Costco accepted an apparent short-term loss on a visible product and fixed the cost structure. IBM used the balance sheet to keep the print looking positive while the franchise thinned. Same word. Opposite sequence.
In my last article, I showed how management writing is not short of data. It is short of the data that would tell a reader whether an idea works where it counts: on a significant scale, in firms that, over a full cycle, perform above average. That test needs to applied to any claim about performance.
Costco and IBM are exhibits, not a proof. The argument needs to go further with a comparison group of firms. When the 30 current DJIA firms and the firms in the STOXX Europe 50 are classified by whether customer value or the quarterly return comes first under stress, the split is crude but stubborn.
Roughly one third operate closer to Costco's sequence. Roughly two thirds operate closer to IBM's. The customer-first group has, on average, delivered far stronger 10-year total shareholder returns. Classification involves judgment. The association is directional, not a scientific proof (which is in any event impossible). The consistency across two major indexes is what makes the pattern hard to wave away.
Figure 1. Dow Jones Industrial Average: ten-year total shareholder return versus customer-focus score. Axis runs 5 (print first) on the left to 1 (customer first) on the right. Nvidia is plotted at a notional +1,500 percent; published decade return is +14,063 percent. Association is directional, not a law.
Figure 2. STOXX Europe 50 on the same axes. Rheinmetall is plotted at a notional +1,500 percent; published decade return is +2,146 percent. Score 2 = customer leans; score 5 = print wins. No firm in this set scored a 1.
That is the firm-level evidence across entire economies that management writing usually skips. It's not a five-step ritual and not a cherry-picked sample. It's the realized performance over a full decade among the most watched public companies in the United States and Europe.
If the returns are so uneven, why do so many famous firms still follow the failing IBM path?
First, a paradigm shift away from traditional management is already well in an advanced stage. Firms built on command, process, and profit-as-purpose do not update the way a spreadsheet updates. The old way of operating is a deeply rooted identity on which many careers still depend. It is tempting to upgrade the vocabulary—"enlightened shareholder value," or "long-term stakeholder capitalism"—while the actual external scoreboard shows no change or decline.
Second, the word "value" has two opposite meanings. For one set of firms, "value" means the customer's situation improved. For the other, "value" means the residual that belongs to shareholders this quarter. Both meanings can sit on the same page of a firm's annual report. They do not describe the same kind of firm. Peter Drucker cut through the fog, back in 1954 in his book, Management: the only valid purpose of a business is to create a customer. Profit is a result and a necessity, not the purpose.
Third, an easy change in wording appears to win. Return on invested capital can be juiced by going slow on innovation. Share buybacks can conceal decline. Three-year relative total shareholder return is not value creation. True value creation is not how the market treats the stock while equity is still vesting.
The 10-year total shareholder return is a public fact, combining backward-looking and forward-looking perspectives. Used with operating evidence, it helps. Used as the maximand of a pay plan, it points to whether the marketplace sees the current roadmap as viable. If not, sanctions can flow, as calls for new management, for splitting the firm, or a takeover. If yes, the firm is awarded support and more resources.
Fourth, the advice industry often stays silent when clients's viewoints diverge from common sense. Last week's article found that pattern in two issues of Harvard Business Review: large claims, thin data support. Bain's call to "bring back managing for value" even used Costco as an illustration of shareholder-value maximization. Costco is an awkward exhibit for that argument. Its public logic has always been customer and employee value first, with shareholder returns as the result.
The tell is what is done under stress, not what is said. When the quarter is underperforming, what gets protected—multi-year innovation, or the quarterly return? When cash is plentiful, does it go into building the next capability or into residual extraction on a 3-year clock? If the CEO's largest goal is 3-year relative total shareholder return, we are looking at shareholder value incarnate. If the only numbers that move the firm are inward—revenue, margin, earnings per share—the window on the customer has already become a mirror.
None of this requires a randomized trial of capitalism, like a scientific with-and-without testing of a new medication. It requires defining success in customer and economic terms; showing a comparison group; showing a full cycle, not a start date chosen to flatter the case. Costco and IBM meet that standard as exhibits. The DJIA and the STOXX Europe 50 extend it. Readers can discuss the details of the classification. They should not have to argue with the absence of any classification at all.
This article was originally published on Forbes.com