Editor’s note: This article was originally written in 2017 by John Deighton and updated in May 2026 to reflect the latest developments in digital marketing and media.
- Tension: Record-breaking AI investment coexists with deteriorating basic customer service across nearly every major industry.
- Noise: Headline-grabbing spending figures create the illusion that technological scale automatically translates to customer satisfaction.
- Direct Message: Infrastructure without intent produces spectacle for investors and silence for the people actually buying things.
To learn more about the DM News editorial approach, explore The Direct Message methodology.
A pattern has emerged across sectors that few executives seem willing to discuss openly. Enterprises are posting record quarterly revenues, announcing ambitious AI deployments, and simultaneously watching their customer service ratings slide to historic lows.
The companies that customers interact with most frequently, from airlines and telecoms to fast food chains and retail banks, have spent years automating touchpoints, optimizing costs, and investing in intelligence layers. Yet the most common consumer complaint remains remarkably unchanged: nobody can get a straight answer to a simple question.
Where is my order? Why was I charged twice? Can I speak to someone?
These queries, trivial on their face, represent the exact moments where brand loyalty either solidifies or collapses. And the distance between corporate spending on technology and corporate attention to those moments has never been wider. The figures are staggering. The ambitions are grand. The hold music is still playing.
The widening canyon between capital allocation and customer contact
When a company reports $202 billion in annual revenue and projects continued growth, the financial press treats the number as evidence of institutional health. Shareholders see confirmation that strategy is working. Analysts revise their price targets upward. The logic appears airtight: scale equals success, and success equals satisfaction somewhere along the chain. But that logic contains a structural blind spot. Revenue measures the ability to extract value from a market. It says nothing about the quality of the relationship through which that extraction occurs.
That imbalance becomes clearer when viewed against the broader marketing economy. In the original archived version of this article, written by John Deighton, data-driven marketing spend was described as accounting for just under 20% of the roughly $1.3 trillion spent annually on marketing in the United States. Even years later, the observation still feels relevant: enormous sums continue flowing into targeting, optimization, and predictive infrastructure while the basic mechanics of customer interaction often remain fragmented and frustrating.
Consider the spending context. As Shira Ovide, technology columnist, has observed, “The $700 billion AI spending spree has few precedents.” Companies across the Fortune 500 are racing to embed generative models, predictive analytics, and automated decision systems into every layer of operations. The appetite for AI infrastructure is enormous and accelerating. Yet much of this capital flows toward internal efficiency gains, advertising optimization, and supply chain refinement, rather than toward the moments where a confused customer picks up the phone or opens a chat window.
This creates a paradox that sits at the heart of modern corporate strategy. The same organizations deploying sophisticated AI to predict purchasing behavior and micro-target ad impressions still route basic billing inquiries through labyrinthine phone trees. The intelligence exists to anticipate what a consumer might buy next Tuesday afternoon, but the systems cannot reliably tell that same consumer why a $14.99 charge appeared on their statement yesterday. The investment thesis privileges prediction over resolution, acquisition over retention, the future customer over the present one.
The consequences are quantifiable. An analysis by the Qualtrics XM Institute found that negative customer experiences are costing businesses $3.8 trillion in global sales annually. Consumers prove most likely to stop or decrease spending with fast food restaurants (66%) and auto dealers (23%) after a bad experience. These figures suggest that the leakage from neglected service moments dwarfs the projected returns from many AI investments. Capital is flowing toward building systems that generate demand while the infrastructure for honoring demand quietly degrades.
When spending headlines substitute for service reality
The noise around AI spending operates as a kind of corporate narrative shield. Every announcement of a billion-dollar compute cluster or a new generative model partnership generates press coverage that frames the company as forward-looking, innovative, and customer-centric by implication. The assumption embedded in this coverage is that technology spending and customer benefit exist in a direct, causal relationship. Spend enough, and the benefits trickle down to the person waiting in the chat queue.
This assumption deserves scrutiny. As Jon Markman has noted, “The $725 billion AI spending surge is missing the real bottleneck.” The bottleneck, in many cases, has less to do with compute power or model sophistication and more to do with organizational will. Routing AI toward customer-facing resolution requires different incentive structures, different metrics, and different definitions of success than routing AI toward cost reduction or ad targeting. The former demands empathy engineering. The latter demands efficiency engineering. Most boardrooms are fluent in one of these languages and functionally illiterate in the other.
Media coverage of quarterly earnings reinforces this imbalance. A company that announces a new AI partnership receives breathless coverage. A company that reduces average customer resolution time from 47 minutes to 12 minutes receives none. The attention economy within business journalism mirrors the attention economy within corporations: flashy infrastructure investments generate narrative capital, while incremental service improvements remain invisible. This creates a feedback loop in which executives allocate resources toward the activities that generate the most favorable press cycles, not the activities that generate the most favorable customer outcomes.
The trend cycle around AI compounds the distortion. Every few months, a new capability, whether multimodal reasoning, autonomous agents, or real-time video generation, captures the industry’s imagination and redirects strategic conversations. Companies feel compelled to announce plans around each new wave, regardless of whether those capabilities address their most pressing customer pain points. The result is a perpetual forward lean toward the next technological frontier, even as the current technological deployment remains half-finished and unevenly applied to the people who actually pay the bills.
The overlooked equation between intent and impact
Research published in the Journal of Travel Research offers a clarifying lens. The study found that intentional service failures, such as overbooking or systematic overcharging, produce far greater negative word of mouth and patronage reduction than unintentional ones. Customers, in other words, can distinguish between a system that fails because it broke and a system that fails because nobody prioritized fixing it. When service gaps persist quarter after quarter while capital expenditure announcements grow quarter after quarter, consumers draw a reasonable inference: the failure is a choice.
The question that separates companies built to last from companies built to impress is whether the intelligence they purchase serves the people who purchase from them.
Redirecting intelligence toward the moments that matter
The corrective here does not require abandoning AI investment or retreating from technological ambition. The scale of spending itself signals something real about the transformative potential of these tools. The issue lies in directional emphasis, in the organizational habit of treating customer resolution as a cost center and customer acquisition as a growth engine.
Several structural shifts would alter this dynamic. The first involves measurement. When executive compensation ties to customer effort scores and first-contact resolution rates with the same rigor it ties to revenue growth and operating margin, resource allocation follows. Metrics shape behavior, and most companies still measure the sophistication of their technology stack rather than the simplicity of a customer’s path to resolution.
The second involves design philosophy. AI systems built for customer interaction require a fundamentally different architecture than AI systems built for internal analytics. They need to handle ambiguity gracefully, escalate to human agents without friction, and prioritize the customer’s sense of being heard over the system’s ability to categorize the inquiry. Many current deployments optimize for deflection, routing customers away from expensive human contact rather than toward effective answers. The short-term savings look compelling on a quarterly report. The long-term erosion shows up in the Qualtrics data.
The third involves narrative honesty. Companies that acknowledge the gap between their technological ambitions and their current service reality build a kind of credibility that press releases cannot manufacture. Customers respond to transparency about known problems and concrete timelines for resolution. The alternative, spending billions on intelligence while the simplest questions go unanswered, produces a dissonance that no earnings call can resolve.
The $202 billion figure, and the hundreds of billions flowing into AI infrastructure across the industry, represents an extraordinary concentration of resources and intention. The question worth asking is whether that intention flows toward the moments that define a customer’s actual experience or whether it flows toward the metrics that define a shareholder’s quarterly satisfaction. These two directions occasionally align. More often than the current spending patterns suggest, they diverge entirely. And in that divergence lives the space where brand loyalty quietly dies, one unanswered question at a time.
