what owning the miss looks like

The sentence that stopped me in Arvind Krishna's letter to IBM investors was only two words: "we faltered." I have seen plenty of executives use similar language while keeping the explanation safely outside their own decisions. Krishna tied his sentence to IBM's judgment, saying the company underestimated the spending shift in front of it and did not adapt quickly enough. Investors now have a management failure they can revisit when the company reports again.

I came to the letter through a Business Insider piece about Krishna taking responsibility for a disappointing quarter. The commentary around the story focused on how unusual the language sounded. I went to IBM's original letter because the useful question was whether the explanation gave people anything they could check later.

Krishna described a market moving underneath IBM. Clients shifted spending toward infrastructure that was already supply-constrained, cybersecurity concerns complicated buying decisions, and some large deals slipped past the quarter. He also described the part that belonged to management: IBM underestimated the size of the change and moved too slowly. The admission placed part of the quarter's cause inside the company.

the market did not wait for the repair

The market response is too large to leave in the background. Yahoo Finance reported that IBM shares closed down more than 25 percent after the preliminary results. Barron's reported that the stock fell another 2.7 percent the next session and estimated that the two-day slide erased more than $72 billion in market value. The five-day chart I pulled late Thursday morning was still down about 28 percent.

I treat that move as behavior under uncertainty. Investors had learned that IBM missed expectations after misreading customer spending and failing to close large deals. They did not need to settle whether Krishna sounded sincere before changing what they were willing to pay for IBM's future earnings.

Owning the miss improved the quality of the explanation, but it did not cushion the consequence. The price action gives us a hard boundary for the trust read. Candor can make management's account more credible while the financial surprise makes management's forecast less credible. IBM now has to repair the second problem with operating evidence.

IBM has more than one trust record

A company this old cannot be read from one quarter or one audience. IBM has a public reputation carried by people who may never sit in one of its product meetings. It also has a quieter record with the enterprises that renew software, build around Red Hat, and keep IBM systems inside operations where failure is expensive. Each group knows a different version of the company because the relationship asks IBM to do a different job.

The public record has held up better than the usual story about IBM fading from view. In 2022, the Axios Harris Poll 100 ranked IBM 11th among the country's most visible companies. YouGov's current U.S. profile reports 85 percent fame, with 50 percent of adults holding a positive opinion, 30 percent neutral, and 5 percent negative. YouGov's underlying quarterly tracker uses roughly 900 to 1,900 U.S. adults per wave.

I would not draw a straight trend line between Harris and YouGov because they ask different questions with different samples. Harris scores the reputation of visible companies among people familiar with them, while YouGov separates awareness from stated opinion. The discipline in our trust work starts here: keep the relationship and the instrument attached to every number instead of blending them into a general halo score.

Interbrand's 2025 valuation gives us another kind of signal. It placed IBM 22nd globally, valued the brand at $39.4 billion, and estimated a 6 percent increase from 2024. Interbrand's estimate is a commercial model rather than a public-opinion poll. I use it as evidence that the IBM name still carries economic weight, with the method boundary kept in view.

enterprise buying has its own clock

IBM's product record has been moving on a longer timetable. The company spent about $34 billion to acquire Red Hat in 2019 and promised to preserve Red Hat's independence and neutrality while building an open hybrid-cloud business. That promise asked enterprise customers to trust two things at once: IBM's scale and Red Hat's freedom to keep working across competing environments.

By 2025, IBM's annual report filed with the SEC showed $67.5 billion in revenue and 6 percent growth at constant currency. Software had reached about 45 percent of total revenue, and IBM reported a generative-AI book of business above $12.5 billion. Those are management's own figures, so I read them as evidence of commercial activity with the company's framing still attached.

Independent reporting captured the mixed signal inside those results. Reuters reported that third-quarter 2025 infrastructure revenue rose 17 percent to $3.56 billion as financial companies adopted the new AI-enabled mainframe. Red Hat growth slowed from 16 percent to 14 percent, and IBM shares fell about 5 percent after hours even though total revenue beat the analyst consensus reported by LSEG. The same quarter held strong product demand and investor disappointment.

Corporate buying gives us behavioral evidence, although it cannot carry the whole trust judgment. A renewal can reflect confidence in performance and the cost of moving a mission-critical system at the same time. The trust model has to leave room for both, because reliance and affection are very different relationships.

the signals can pull in different directions

Our circumplex work assumes that people can value different, sometimes opposing, qualities and still be making legitimate trust judgments. An enterprise buyer may lean toward continuity and control, while a developer puts more weight on openness and freedom to move. IBM's Red Hat promise sits close to that tension, asking the market to believe the combined company can provide both without hollowing out either side.

The public respondent occupies another pathway because many of the adults answering YouGov know IBM as a long-lived technology name without directly buying its current products. Their response tells us how the company sits in public memory and present opinion. It does not answer whether a bank trusts a new mainframe, just as a Red Hat renewal does not tell us whether the public feels close to IBM.

Investor confidence runs on a faster clock because the market can believe IBM has durable enterprise relationships and still punish a forecast miss when the expected growth story breaks. The selloff records a change in expectations around management performance, with no need to pretend every other form of trust moved by the same amount on the same day.

This source stack is how I want to read a company such as IBM over time. Public opinion and product revenue describe different relationships: one records stated reputation, while the other shows where organizations keep committing money. Market response adds a faster measure of confidence in leadership's forecast. Keeping the disagreements visible makes the sources more useful.

the admission now has a longer context

Krishna's letter sits inside a company with accumulated public credit and deep enterprise reliance. A bad quarter does not erase either record, although it can expose a weakness in management's ability to read change across the products and customers supporting them, which is the failure Krishna chose to name.

The preliminary results also show why a simple trust-up or trust-down score would miss the story. Customers were still spending, but their priorities moved toward supply-constrained infrastructure as prices were expected to rise. IBM's miss came partly from where that money moved and partly from management's slow response. The admission earns attention because it separates the market condition from the execution failure without hiding behind the first one.

IBM's full earnings call is scheduled for July 22. I will listen for how those customer signals entered the forecast and what changed after leaders saw large deals slip. A later quarter can show whether the adjustment held. Repeated updates will show whether management keeps the same clarity after the immediate pressure fades.

I am less interested now in praising one candid sentence. The stronger read is that IBM has accumulated several kinds of trust over time, and this quarter exposed a gap between customer behavior and management's forecast. Krishna put management inside that gap, leaving IBM's next evidence to show whether the company learned how to see it sooner.

The next post moves from investor trust to operational harm. Telstra and Capita apologized after failures had already reached people who depended on their systems, raising a harder question about the evidence an apology owes the people carrying the consequences.

sources

Next in the series: an apology needs a repair ledger.