// attention economy

All signals tagged with this topic

McKinsey finds AI productivity gains depend on organizational execution

McKinsey's latest research confirms that AI delivers measurable efficiency improvements—but only for companies that actually restructure workflows around it rather than bolting it onto existing processes. Organizations seeing outsized gains are actively eliminating redundant roles and retraining staff, while those treating AI as a plug-and-play tool watch returns flatten. For consumer-facing businesses, this means competitive advantage goes to companies disciplined enough to make unpopular operational changes, not simply those with the biggest AI budgets.

Word-of-Mouth Still Drives Podcast Discovery, Not Algorithms

The podcast industry's growth narrative obscures a stubborn distribution reality: despite years of investment in AI curation and algorithmic recommendation, listeners still predominantly discover shows through social signals rather than platform discovery tools. This disadvantages high-volume "slop" producers who bet on algorithmic amplification, forcing them to compete on content quality and cultural relevance instead—a different economics than YouTube or TikTok, where algorithmic serendipity can drive engagement regardless of word-of-mouth merit. The gap between podcast production capacity and actual listening attention is widening because the medium hasn't solved the discovery problem that would unlock passive consumption at scale.

OpenAI's Stalled Revenue Growth Exposes Consumer AI's Monetization Problem

OpenAI's revenue figures show a gap between ChatGPT's 200 million monthly users and actual paying customers. The mismatch suggests that free trials and freemium models have trained users to treat AI as a commodity utility rather than a premium service. The company now faces pressure to prove that chatbots and generative interfaces can sustain venture-scale economics through subscription revenue alone. The shortfall could redirect the $50+ billion in AI investment away from consumer subscriptions toward enterprise licensing and infrastructure, where customers have measurable budget constraints and ROI requirements.

Why AI Art Actually Wins With Consumers

The article inverts the typical quality-first assumption: people don't choose AI outputs despite inferior technical craft, but because they actively prefer the aesthetic and emotional register of lower-friction, less-perfected work. This preference isn't a flaw in consumer taste—it's an advantage for generative tools over traditional art production, where the polish that institutions and gatekeepers have trained us to value becomes a liability in markets that reward novelty, personalization, and the uncanny. The competitive threat to human creators isn't AI matching their skill; it's AI matching what audiences actually want, which is often the opposite of what art schools teach.

Google's AI Overviews Are Surfacing Negative Reviews Unprompted

Google's AI Overviews are displaying critical reviews to users without triggering searches for complaints—a shift in how negative sentiment reaches consumers at the moment of purchase consideration. This creates a new vulnerability for brands: instead of controlling narrative through review management and SEO, companies now face algorithmic curation that prioritizes relevance over sentiment, potentially amplifying criticism that users never explicitly sought. For e-commerce and service businesses, reputation management must evolve beyond review suppression tactics to actual product and service quality, since the algorithmic middleman can surface problems regardless of search intent.

Why BlackBerry's Keyboard Obsession Blinded Them to the Touchscreen Future

Seth Godin's post examines how BlackBerry executives confused vocal user preference with market demand, doubling down on physical keyboards while the market shifted toward touchscreens. The trap: mistaking the loudest existing customers (enterprise users adapted to keyboards) for the broader market appetite. That bias cost BlackBerry its dominance to Apple and Android devices. For consumer product makers today, the risk is treating feedback from your most entrenched users as a proxy for where consumption is actually moving.

Smartphone upgrade cycles stretch to 4.2 years as inflation bites

Consumers are extending device lifecycles in response to economic pressure. The average phone now lasts nearly a year longer than a decade ago, and handset manufacturers are operating in a structurally lower-velocity replacement market. This shifts competition toward durability and repairability rather than planned obsolescence, while strengthening secondary markets for refurbished devices and independent repair services that incumbents have historically suppressed. For hardware makers, fewer upgrade cycles compress revenue directly, making software services, subscription models, and ecosystem lock-in increasingly critical to survival.

Roblox's Age Verification Push Backfires on User Growth

Roblox is caught in a classic platform trap: the safety measures required to attract older users and advertisers (age checks, content moderation) actively repel the younger audiences who built the ecosystem in the first place. The company's DAU decline shows that compliance isn't costless—platforms can't simply bolt on gatekeeping without friction bleeding into retention. Roblox faces a choice between regulatory legitimacy and the network effects that drive valuations. Metaverse platforms will confront the same tension as they navigate COPPA enforcement and advertiser demands simultaneously.

India Becomes ChatGPT's Image Generation Beachhead

OpenAI's image generation tool is finding its earliest and strongest adoption in India, where users deploy it for practical creative work—avatars, portraits, design assets—rather than novelty use. This geographic concentration reflects a straightforward economic pattern: generative image tools gain traction first in markets with high creative labor costs and limited access to design software, not in saturated Western markets where Midjourney, Adobe, and others already provide similar capabilities. The adoption gap between India and the West shows that AI uptake follows economic logic: the tool becomes essential where it solves a real scarcity problem.

AI Therapy Apps Show Promise, But Evidence Still Sketchy

The mental health startup ecosystem has built a multi-billion dollar business on the assumption that algorithmic chatbots can meaningfully substitute for human clinicians—a bet that depends entirely on clinical outcomes we still don't have at scale. New research suggesting mixed or modest efficacy challenges the narrative that AI can democratize therapy, forcing apps like Woebot and Replika to either generate harder evidence or position themselves as supplements rather than alternatives to licensed care. Venture capital will follow the data: if randomized trials consistently show weak effects, funding dries up and the market consolidates around players who can survive on lower user expectations and smaller revenue bases.

Meta Cuts Ties With Labelers After Exposing AR Glasses Data Pipeline

Meta terminated its contract with Sama, a Kenyan data labeling firm, after Swedish journalists revealed that low-wage workers were reviewing sensitive video feeds from Meta's Ray-Ban smart glasses. The move exposes the company's reliance on human annotation for AI training and the vulnerability of workers who see proprietary product data before public release. The termination amounts to damage control rather than an ethical reckoning. Meta's public commitment to responsible AI development depends on a precarious chain of contractors in the Global South who lack legal protections and can be dismissed once their labor becomes a liability. The pattern runs deeper than one vendor relationship. As consumer AI products embed cameras and sensors into everyday objects, the human work required to manage the gap between technical capability and acceptable privacy standards moves offshore and out of sight.

AI Widens the Productivity Gap Between Junior and Senior Workers

An MIT study of 5,000 customer-service agents found that generative AI boosted novice workers' productivity by 34%. The shift is structural. Junior employees now access institutional knowledge and problem-solving support that previously required years of mentorship or peer networks. Senior workers can no longer rely on experience gatekeeping as a competitive advantage. The economic pressure flows upward: companies adopting AI assistance for entry-level roles immediately question why they're paying for expensive senior talent when less experienced workers, augmented by LLMs, can close that gap in months rather than years.