One of the things that I’ve been following closely over the years has been the demise (and the continuing difficulty faced by) casual restaurants, for a variety of reasons. All the places we used to go out when I was young – Olive Garden, TGI Fridays, Chili’s, Pizza Hut – they’re slowly giving way. 

Some of that, of course, is the rise of the delivery services (DoorDash, etc.) and the segment of Millennials and Gen Z who don’t really enjoy going out to eat.

But it’s not just that reason, and it’s not just chains.

Many mom & pop restaurants around the country serving that same segment of the population face similar headwinds: rising costs, lower foot traffic, and an overall general souring of the economic mood.

Indeed, the restaurant industry is part of a larger trend we’re seeing, one tied to the fact that wealth has been increasingly concentrated at the top, while we’ve hollowed out the middle class. And this “middle,” we believe, is going to face extreme challenges over the next several years.

Brands Are Designing Only for ‘The Wealthy’

In the midst of waiting for my American Airlines flight (which, for the second time in a month, doesn’t have enough staff on hand to fly its scheduled 5:38 p.m. to Cleveland), I was reading about the airline’s recent financial projections. The airline’s new focus: premium offerings for high-spending travelers (while clearly ignoring basic staffing for ordinary flights). You can read about my previous customer experience here.

It’s a story being told across multiple industries today. The middle of the consumer economy is disappearing. For decades, brands emphasized products and experiences around a broad middle class: a reasonably comfortable flight, a clean mid‑tier hotel, a family trip to Disney, a night out at a concert that didn’t require months of saving.

Today, more and more markets are splitting into two worlds: ultra‑premium offerings tailored to the affluent, and stripped‑down, bare‑bones options for the price‑sensitive, with very little in between. The connective tissue is data and AI: granular segmentation, personalization, and dynamic pricing that optimize for maximum willingness to pay rather than for a shared “average” experience.

Companies used to talk about their “typical customer” or “average family.” Now they talk about cohorts, micro‑segments, and 1:1 personalization. That language shift signals something deeper: once you can see every customer as a unique revenue curve and adjust the product and the price accordingly, the economic incentive to maintain a robust, fairly‑priced middle tier weakens. Instead, you push the top to pay more for more, and press the bottom into accepting less for less.

“Average,” in other words, is being optimized out of existence.

Airlines: From Mass Transit to Micro‑Markets

For a long time, commercial flying was a relatively simple hierarchy: economy, business, first. The vast majority of travelers flew economy, and while there were differences in comfort, the basic experience was recognizable: a seat with tolerable legroom, a checked bag, and some sort of meal on longer flights. This was air travel as mass transit for the middle class, not a luxury good reserved only for elites.

Over the last two decades, airlines have systematically dismantled that model. Cabin structures have proliferated into basic economy, standard economy, extra‑legroom economy, premium economy, business, first, and fully enclosed suites. Each of these is further sliced by restrictions on changes, refunds, and miles earning. Academic work on airline pricing shows that this isn’t random complexity: it’s a deliberate, product‑based price discrimination designed to capture more of each traveler’s willingness to pay. The old, broadly accessible “middle” product (comfortable, all‑in economy) shrinks, while both extremes expand: rock‑bottom basic fares at one end, lavish suites at the other.

AI supercharges this logic.

Modern revenue‑management systems ingest live booking data, competitor fares, historical demand curves, and customer signals to adjust prices and fare availability in real time.

It’s like a surge-pricing model if you’ve encountered these recently in other realms. For example, I was about to take an Uber to the airport this morning. To go roughly 3 miles, the company wanted to charge me $55 because it was a period of high-demand and it concluded I could probably pay that. (I ended up taking a Lyft, which doesn’t have much of a payment history for me, for almost half the price.) 

For airlines, machine‑learning models classify travelers by how far in advance they book, how often they fly, what cabins they’ve paid for before, even what devices they use, and push them toward a highly profitable option: either a bare‑bones basic ticket plus fees or an upsold product like premium economy or business. The standard “middle” economy ticket quietly morphs into a moving target: sometimes scarce, often priced closer to what used to feel like a premium fare.

For a middle‑class family of four, the practical choice begins to look like this: endure a stripped‑down, fee‑ridden basic experience: no seat selection (which happened to me on a recent flight), last to board, risk of getting split up, or pay hundreds more just to restore what, 15 years ago, was considered normal. AI is doing precisely what it’s designed to do: what economists call surplus extraction, and we recently termed the new “extraction” economy.

The collateral damage is the disappearance of a shared, baseline level of comfort that defined the mainstream flying experience.

Lodging: The Vanishing Mid‑Market Hotel

Drive across the United States and you’ll notice something similar in hospitality. For decades, the country was knit together by a carpet of midscale hotels and motels: Holiday Inns, roadside motor lodges, basic but predictable chains that provided clean rooms for families, businessmen, or other tourists depending on the locale. They were not glamorous, but they used to be the backbone of middle‑class travel.

Industry analyses now describe a “hollowing out” of this segment. In major cities and desirable leisure markets, older midscale properties are being torn down, repositioned, or converted into upscale apartments and boutique hotels. New construction disproportionately flows into two buckets: high‑end lifestyle and luxury brands on one side, ultra‑budget motels on the other. The mid‑range (those reliable, modestly priced, full‑service hotels with decent space and amenities) has become far more fragile, especially in markets where land values and debt costs push owners to chase higher‑yield segments.

Behind the scenes, hotel groups deploy sophisticated revenue‑management tools that dynamically adjust room rates by night, room type, and demand forecast. When events, seasons, or compression hit, these systems push nightly prices of remaining decent mid‑scale rooms up toward luxury levels, especially when midscale supply is constrained. AI‑driven demand models make it rational to squeeze more out of the few “average” rooms that remain.

At the same time, luxury properties layer on club floors, suites with exclusive lounges, and tailored offers for top‑spending guests driven by AI‑powered CRM and segmentation. Budget hotels strip services down to the bone: minimal staff, no breakfast, smaller rooms, often in less desirable locations. The traveling middle class, once assured of a broad middle band of comfortable, affordable options, is nudged toward either paying up for quasi‑luxury or settling for an experience that looks and feels distinctly below the standards they grew up with.

Disney World Is Out of Reach: Magic Only for the Few

If airlines and hotels show how the journey has been transformed, Disney shows how the destination has changed. Walt Disney World used to be the quintessential middle‑class aspirational trip, a once‑every‑few‑years splurge that was expensive but achievable.

But costs have gotten so out of hand that it’s not an absurd idea to fly a family to Tokyo Disneyland to save money, depending on what part of the U.S. you’re coming from.

How bad is it?

Single‑day tickets for adults frequently land in the 150–200‑dollar range on many dates, and multi‑day passes stack into the high hundreds per person. Meanwhile, perks that were once bundled into the ticket (FastPass, broad extra‑hours access) have been spun out into paid add‑ons like Lightning Lane, after‑hours parties, and special events.

Disney has also adopted increasingly granular date‑based pricing and is moving toward dynamic pricing models that behave more like airlines: higher prices on high‑demand days, lower on off‑peak dates, with the system constantly tuned to demand patterns. That may be efficient from a revenue perspective, but it penalizes families tied to school calendars and fixed vacation windows. You know, the classic middle‑class households without flexible jobs or year‑round travel freedom. Wealthier visitors, by contrast, can choose slower weeks, pay for line‑skipping and VIP tours, and stack premium experiences on top of their base tickets.

The global comparison sharpens the point. Tokyo Disney Resort’s published ticket prices are substantially lower than those at U.S. parks, often in the 50–70‑dollar range per adult per day, with kids’ tickets lower still. Trip‑cost breakdowns by travel bloggers and analysts have found that a six‑night Tokyo Disney vacation, including mid‑range lodging and multi‑day park tickets, can sometimes undercut, or at least closely match, a comparable Walt Disney World trip. This is especially true for travelers who can access competitive fares to Tokyo from major U.S. hubs, because in‑park costs, local transportation, and food are often cheaper in Japan than in Orlando.

Think about it: For some Americans with decent incomes and good airport options, it can be economically rational to fly halfway around the world rather than drive or fly to Florida to experience roughly similar Disney magic. That inversion is not just about exchange rates; it’s about Disney World’s steady migration up‑market, enabled by dynamic pricing, elaborate upsells, and a strategic embrace of high‑spend guests. Also, the fact of the matter is that the Japanese don’t simply try to EXTRACT EVERY DOLLAR FROM YOU, while we in the U.S. have hesitation to do so.

In effect, it seems crazy to say this, but Disney World in the U.S. is a luxury resort brand.

Live Events: The VIP‑ification of Fun

Live events—concerts, sports, theater—used to be the social glue of the middle class. I sound like a broken record, but back in the ’80s or 90s even, you might not sit on the 50‑yard line or in the front row, but you could afford a decent seat without wrecking the household budget.

Those days are gone.

Data on concert economics show that average ticket prices have increased more than tenfold since the 1980s. In 2024, the average ticket price hovered around $136 (and you’re not getting that close to anything at that rate). In 1985, we went for $12. Literally.[1]

In practice, for big tours and major league games, mid‑level seats often land in the 120–280‑dollar range after fees, even without any VIP or meet‑and‑greet extras. For many families, especially once you add parking, concessions, and merch, that transforms a simple night out into something that looks more like a luxury expense.

Some of you, I know, don’t have to imagine taking a family of four to a sporting event for $1,000. Because that’s the price of a decent seat, food and drinks.  

Here, too, algorithms lurk in the background. Ticketing platforms rely on dynamic pricing and “platinum” tiers that raise prices in response to demand surges, using real‑time sales patterns and search data to find the price ceiling. AI‑driven forecasting tools help promoters set initial prices and adjust them based on everything from social buzz to day‑of sales, ensuring that the hottest seats for the hottest acts capture as much high‑income demand as possible. The stadium or arena becomes a visible map of stratification: glass‑walled clubs and field‑level suites for the top, a thinning stripe of acceptable mid‑tier seats in the middle, and packed nosebleeds at the top of the bowl for everyone else.

What’s being hollowed out is not only affordability but equality of experience. The modern venue is built to deliver fundamentally different worlds to different income tiers, from private entrances and gourmet dining to cramped bleachers and long lines. AI and dynamic pricing ensure that each segment is charged as much as the market will bear.

Retail and E‑Commerce: When the “Average Shopper” Disappears

Retail is where the logic of AI‑driven personalization becomes most explicit. For much of the twentieth century, the core tools of retail marketing were broad strokes: weekly circulars, mass coupons, seasonal sales. Everyone saw roughly the same prices, the same promotions, the same end‑caps in store. The archetype was the “average shopper,” a composite persona used to plan assortments and layouts.

AI‑powered segmentation has rendered that archetype obsolete. Modern e‑commerce and retail marketing platforms champion 1:1 personalization, driven by clustering algorithms and behavioral data that group shoppers by their browsing, purchasing, and response patterns.

Dynamic pricing tools can adjust product prices, discounts, and promotional offers in near real time as conditions change. In many cases, this doesn’t yet mean that two people see different numeric prices for the exact same item at the same moment, but it does mean that they see different product mixes, different discount levels, and different nudges.

The distributional consequences are complex. High‑value, high‑spend customers may receive better service, early access to limited items, and loyalty perks, but research on personalized pricing suggests they can also be quietly steered toward higher‑priced options and face fewer discounts because their willingness to pay is higher.

Lower‑income, more price‑sensitive customers may get more coupons or participate in more gamified promotions, but they are also more exposed to dark patterns: interfaces designed to manipulate choices, and to product assortments that prioritize margin over quality.

What’s missing in this environment is a stable, widely understood “fair” price. The concept of a shared middle ground (what a loaf of bread or a pair of jeans costs for everyone), we believe will be eroded by granular targeting. Every shopper becomes a separate optimization problem.

The average, hence, is not a person; it’s an output of a model.

AI Is Set to Wipe Out ‘Average’

If you’ve used the tools at all over the last several years, you realize one thing: Generative AI, in particular, is getting very good. It’s no longer a novelty that just writes a somewhat awkward, but interesting passage or creates an image that comes close to resembling something. Today:

• • It writes better than the average person.
  • It designs better than an average designer
  • It creates compelling videos better than an average filmmaker and videos that trick people online every day, who don’t realize it’s not real
  • It does voiceovers that are so good you can’t tell it’s not human
  • It researches faster and more effectively than any human being or team of human beings can do.
  • It drives cars more safely than ordinary human beings (outside of strange hiccups when the power goes out)
  • It knows more legal and case history than any single lawyer or team of lawyers
  • It can identify some diseases or other issues better than an average doctor

The list can go on and on.

Here’s the cold reality that we believe is coming soon: If you do average work, AI will very soon replace that work.  

Organizations with capital and skills can deploy AI for automation, analytics, and personalized services, boosting productivity and profitability. For an interim period, companies will need people to help implement that AI. But after a while, they won’t.  

In other words, AI is not just carving out the middle in consumer markets; it is also carving out the middle in opportunity. The same tools that let a retailer identify its most valuable customers let an employer identify its most “efficient” workers, and potentially its most replaceable.

For the average person, we’ve talked a lot in this blog about the importance of getting really good at understanding other people and the different nuances that machines can’t pick up (because they don’t really understand people).

In other words, if you’re good with other people, you will likely give yourself the best opportunity to continue to work and survive, because that quality above all others will still be valued, in our opinion.

Conclusion: The Barbell Economy Is Already Here: Haves, Have‑Nots, and the Hollow Middle

If you step back from the sector details, a consistent structure emerges.

At the top of the barbell are affluent consumers and businesses. They experience AI as a concierge and a lubricant:

• • Priority boarding and lie-flat seats on planes.
  • Club-level rooms and personalized service in hotels.
  • VIP tours and line skipping at theme parks.
  • Front row concert seats and exclusive lounges.

At the bottom are the price‑sensitive and the digitally disadvantaged. They experience AI as a gatekeeper and a rationing system:

• • Basic economy with nickel and diming and uncomfortable seats.
  • Budget hotels with inconsistent quality and minimal services.
  • Long lines at theme parks, with key conveniences paywalled.
  • Nosebleed event seats, if they go at all.

In between lies a shrinking middle: the family that can still afford to fly but agonizes over seat fees; the traveler who can manage one nice hotel night but not a full week. The Disney‑loving parents comparing spreadsheets to decide whether it’s worth going at all. The fan who skips the concert and watches clips online. Their experiences are increasingly defined by trade‑offs and trade‑downs, not by the confident expectation that the system is designed with them in mind.

AI doesn’t create inequality, but it makes inequality more legible and more actionable.

Think about it: What happens to social cohesion when the notion of a shared “normal” experience (whether on a plane, in a hotel, at a stadium, or even in front of a TV) fractures into one world for those whom AI serves, and another for those whom AI manages or simply takes over?

Those are not abstract questions.

They show up in the family that drives instead of flies, the fan who watches the game from home because the stadium has become a luxury venue. Piece by piece, the “average” is being engineered out of the system. The open challenge is whether all of us can recognize this now and protect a robust middle in an era where the math of AI makes extremes so tempting.

[1] You can read more about this in our Extraction Economy piece: https://marketingniceguys.com/welcome-to-the-extraction-economy/