The contemporary consumer landscape has acquired a remarkably narrow palette. Cars are overwhelmingly white, black or grey. New apartment buildings favour muted façades and interchangeable materials. Consumer electronics have largely abandoned the coloured plastics that were common even twenty or thirty years ago. Restaurants, hotels and renovated homes increasingly circulate through the same combinations of beige, pale wood, off-white, black metal and desaturated green.

It is tempting to treat this as another complaint about minimalism, or as a form of nostalgia in which the past appears more colourful simply because memory retains its most distinctive objects. There is some truth in both explanations. Design fashions change, and historical comparisons are easily distorted by selection.

Yet the loss of colour is also visible in data.

An analysis by the Science Museum Group examined more than 7,000 photographs of familiar objects from 21 categories, including domestic appliances, photographic equipment, clocks, lighting and writing technology. Across the collection, the researchers found a clear increase in grey over time, accompanied by declines in brown and yellow. Their study of telephones was particularly revealing. Phones from the 1960s, 1970s and 1980s appeared in a much broader range of colours, while the shift back towards black, silver and grey began in the late 1980s and continued into the smartphone period. The researchers were careful to point out that a museum collection is not a representative sample of everything people owned, but the pattern was strong enough to be visible across thousands of objects.

Cars provide a larger and cleaner example. An iSeeCars analysis of more than 22 million vehicles found that black, white, grey and silver accounted for 47.3% of the US car market among 1996 model-year vehicles. By 2025, their share had reached 80.4%. Green fell from 13.4% to 2.2%, while red declined from 20.1% to 7%. Grey alone rose from 3.6% to 22.9%.

Housing data suggests that the palette may now be warming without becoming substantially more colourful. In the 2026 US Houzz Outdoor Trends Study, beige became the most popular exterior wall colour, rising from 10% of renovating homeowners in 2024 to 24% in 2026. White and grey both declined. This may mark the retreat of the cool-grey interiors and façades that dominated much of the previous decade, but replacing grey with cream, sand and greige is not exactly a return to chromatic abundance.

The more interesting question is why so many unrelated categories seem to move in a similar direction.

Preference is only part of the market

The obvious answer to the prevalence of neutral colours is that consumers prefer them. If people keep buying grey cars and beige kitchens, manufacturers and developers will keep supplying them.

That explanation is incomplete because supply and preference are not independent.

Consider the process through which a car colour reaches the road. A manufacturer decides which colours to produce, how many colours a model should offer and which combinations will be available with particular specifications. Fleet buyers and leasing companies place large orders. Dealers decide which cars to hold in stock. Buyers often choose among those immediately available rather than waiting months for a factory order. The same car eventually enters a used market in which future resale value becomes part of the original purchasing decision.

At every stage, colour is exposed to a different calculation of risk.

A buyer may personally prefer green but accept grey because the grey car is already available. A fleet manager may prefer colours expected to be easy to resell. A dealer carrying several hundred vehicles has little reason to take a large position on a colour that divides opinion. A manufacturer looking across several markets has an incentive to keep production complexity under control.

The resulting colour distribution is therefore not simply a referendum on what individual consumers would choose if every option were equally available, equally priced and immediately obtainable.

There is an interesting contradiction in the resale argument itself. A separate iSeeCars study of more than 1.2 million three-year-old vehicles found that yellow, orange and green cars actually depreciated less than the market average, partly because their supply was limited relative to demand. Black and white, despite being among the most common colours, depreciated more.

In the study, the average vehicle lost 31% of its value after three years. Yellow lost 24%, orange 24.4% and green 26.3%. Black lost 31.9% and white 32.1%.

This does not mean everyone secretly wants a yellow car. It does show that the assumption that neutral colours are economically safer can become self-reinforcing even when the secondary market does not always support it.

The distinction matters because markets do not merely aggregate preferences, they also decide which preferences are cheap enough, predictable enough and sufficiently scalable to satisfy.

The economics of not being disliked

Suppose one design is strongly preferred by 30% of customers, acceptable to another 30% and disliked by the rest. A second design inspires little enthusiasm but causes almost no objection.

The first has a constituency - but the second has a market.

That difference becomes increasingly important as the intended audience grows.

A local restaurant can survive by appealing intensely to a particular group of people. A hotel chain operating in fifty countries has a different problem. A developer selling hundreds of apartments cannot know the taste of every eventual buyer. A global electronics company has to place the same product into markets with different aesthetic traditions, retail structures and income levels.

Under those conditions, avoiding rejection becomes economically useful.

This does not require companies to conduct formal surveys asking which colour nobody hates. The same effect can emerge through sales data, inventory management, retailer feedback, conversion testing, returns, resale assumptions and accumulated managerial experience. Strongly polarising choices generate visible downside, while broad acceptability is easier to defend.

The result can look like consumer preference even when it is partly the product of repeated filtering.

This is where the spread of neutral design becomes more interesting than a sequence of fashions. The underlying selection mechanism favours characteristics capable of surviving many different audiences and many different decision-makers.

It can produce a world increasingly composed of things that the fewest people object to.

That is not equivalent to producing the things people like most.

Universality and the removal of specificity

Modern consumer markets operate at scales for which most historical design cultures were never intended.

A chair made for one household could reflect the taste of that household. A restaurant designed by its owner could reflect a particular idea of what a restaurant should look like. Regional building traditions emerged from local climate, available materials, craft practices and inherited conventions.

The modern mass-market object is under different pressure. It may have to be manufactured efficiently in several countries, transported through standardised logistics, photographed well for online retail, satisfy a global brand system, fit multiple interior styles and remain commercially viable for years.

Specificity becomes expensive.

Colour is one of the easiest forms of specificity to remove, but the same tendency affects shape, material, typography and ornament. The Science Museum analysis found not only that recent objects had become somewhat greyer, but also somewhat squarer. Modern electronics illustrate both changes particularly well. Earlier technological objects often exposed their components and mechanical structure. Contemporary devices increasingly conceal them inside smooth rectangular shells.

Architecture faces similar pressures at a much larger scale. Building systems, international suppliers, standardised components, investment models and large development groups allow the same architectural vocabulary to travel easily between cities. Research on historic districts increasingly treats colour as part of local cultural information precisely because homogenisation can remove distinctions that once made one urban environment visibly different from another.

A 2026 study of architectural colour across 153 buildings in eight historic districts of Dalian compared four historical periods between 1898 and 2014. The researchers explicitly frame “urban colour homogenisation” as a concern and found that two of the recurring colour groups appearing only after 1979 consisted of low-saturation neutrals.

The evidence here is more local than the car data, and it would be wrong to claim that every city has become objectively greyer by the same measurable amount. The relevant point is that researchers now have reason to study urban colour homogenisation as a phenomenon rather than merely as an aesthetic complaint.

The marketing problem hidden inside the beige room

Marketing has spent much of the last several decades becoming better at reducing uncertainty.

Segmentation reduces uncertainty about the audience, research reduces uncertainty about demand, brand guidelines reduce uncertainty about presentation. A/B testing reduces uncertainty about which message will perform better. Recommendation systems reduce uncertainty about what to show next. Conversion optimisation removes elements that appear to obstruct action.

Individually, these are sensible practices. I use many of them professionally.

The problem appears when the logic of optimisation becomes confused with a theory of what is worth producing.

Metrics are usually much better at detecting resistance than attachment. If a design causes customers to abandon a purchase, complain, return an item or reject a property, the effect is visible quickly. The value of distinctiveness is often slower and harder to isolate.

A restaurant interior that some customers dislike may also be the reason others remember the restaurant five years later. A strange building may perform poorly in a preference survey and nevertheless become a local landmark. A colour that reduces the addressable market may create a much stronger preference among the people who remain.

Average performance can conceal that difference.

Marketing researchers have long known that averages can disguise distinct segments. The practical response is normally better segmentation. But mass production, property development and global branding place limits on how far segmentation can be carried into the physical object itself.

It is cheaper to personalise the advertisement than the physical product.

It is easier to show twenty audience segments different campaigns than to construct twenty versions of an apartment building.

The market can therefore become highly personalised at the level of communication while remaining increasingly standardised at the level of the shared physical environment.

Personalised feeds, standardised surroundings

This produces one of the stranger characteristics of contemporary consumer culture.

Digital environments have never been more individually tailored. Two people opening the same platform may see almost entirely different worlds. Music, news, entertainment, advertisements, shopping suggestions and search results are increasingly selected according to personal behaviour.

Meanwhile, many of the physical objects through which those people encounter the world have moved in the opposite direction.

The smartphone is available in a handful of restrained finishes, but the screen contains a personalised universe. The apartment development uses the same neutral palette across hundreds of units, while the occupants receive individually targeted media. The hotel room could be in several countries, while the traveller's apps know exactly where she is.

Difference has not disappeared. Part of it has migrated from the public environment into the private interface.

That has commercial advantages as software can vary endlessly at very low marginal cost, while physical production cannot. A recommendation feed can present a different arrangement to every user without manufacturing a million different phones.

But there is a cultural consequence to locating more individuality inside screens while removing it from streets, products and buildings.

Shared environments contain less evidence of other people's preferences.

Algospeak has a physical cousin

The mechanism resembles the one behind algospeak, although the outcome is obviously different.

Algorithmically shaped language did not arise because millions of users independently concluded that words such as “unalive” were more expressive than the words they replaced, but because creators learned to anticipate systems whose rules and commercial consequences were difficult to observe. When the possible penalty for using a direct term was reduced reach or monetisation, while the cost of replacing it with a euphemism was relatively small, caution became rational.

No explicit prohibition was required. Repeated adaptation to perceived risk was enough.

Commercial design can operate through a related process. Nobody needs to issue a rule stating that cars should be grey, apartments beige or phones black. Each participant makes a smaller decision about production, inventory, resale, broad appeal or cost.

The accumulated result can be much more uniform than the preferences of any individual participant.

This is an important distinction because explanations based entirely on taste overlook the institutional environment in which taste becomes a product.

People choose, but they choose from menus created by other choices.

The limits of the argument

It would be easy to overstate the case.

There has never been a uniformly colourful past. Early consumer technologies were often black, brown or metallic because of their materials and manufacturing methods. Victorian cities were not psychedelic. Modernist architects were using white, concrete and glass long before contemporary lifestyle brands discovered beige.

Some neutralisation is technological rather than ideological. The Science Museum researchers themselves note that the decline of brown partly reflects the replacement of wood by plastic and other materials.

Nor is colour disappearing in a straight line. Beige is currently displacing some grey in housing, while automotive colour data suggests the long rise of grayscale vehicles may have plateaued at around 80%. Sports cars remain considerably more colourful than the general market: non-grayscale colours account for 36.2% of sports cars, compared with 19.6% across the overall vehicle market.

There are also areas where colour has expanded enormously. Digital displays can produce ranges of colour unavailable to previous generations. Fashion cycles regularly return to strong palettes. Packaging remains intensely competitive for visual attention.

The claim is therefore not that modern life is literally losing colour everywhere, it is narrower, and more interesting: in categories where products must satisfy large and uncertain audiences, pass through several commercial intermediaries or preserve future resale value, there are strong incentives to remove characteristics that create rejection.

Colour is simply one place where the result can be seen.

Pepperland without the Blue Meanies

This is where Yellow Submarine provides an unexpectedly useful metaphor.

When the Blue Meanies take over Pepperland, the damage is represented through the removal of colour, music and absurdity. The landscape still exists, but its emotional content has been stripped out. Colour functions as shorthand for a society in which people and things are allowed to be exuberantly particular.

The comparison should not be pushed too far. There is no equivalent villain making contemporary products beige, and commercial standardisation is not cultural repression.

The unsettling part is almost the opposite.

A great deal of visual uniformity can emerge without anybody intending to create a uniform world.

The property developer wants a finish acceptable to the largest number of buyers. The manufacturer wants a product that can be sold across several markets. The dealer wants inventory that will move quickly. The hotel group wants a design that can be reproduced reliably. The marketer wants the version that performs adequately across the largest audience. Each decision is defensible within its own frame.

However, the cumulative environment is nobody's explicit preference. This is a familiar feature of market systems. Outcomes can emerge from individually rational decisions without corresponding to a collective intention. Economists usually discuss this in relation to prices, congestion, externalities or resource allocation. Aesthetic convergence deserves to be considered in the same way.

The emotional consequence is difficult to price because it appears only in aggregate. One grey car is not depressing. Neither is one beige hotel, one black phone or one pale apartment block. The effect comes from repetition.

Human attachment is unusually dependent on distinction. We remember that strange café, the coloured telephone in a grandparent's house, the particular tiled façade, the car colour that immediately identifies a period. Objects become mnemonic partly because they are not interchangeable.

An environment selected primarily for broad acceptability risks losing some of that capacity. It may become easier to sell and harder to remember.

The least objectionable world

Marketing usually treats broad appeal as an achievement, and often it truly is. But broad appeal can describe two different things.

A product may contain qualities that many people positively value. Or it may have been stripped of qualities that significant numbers of people might reject.

Those routes can produce similar survey results while producing very different culture.

The difference matters increasingly because the systems making commercial decisions are becoming better at identifying resistance. Digital testing can detect small conversion losses and platforms can compare thousands of variants. AI can generate endless alternatives and select them against performance criteria. Product teams can quantify behaviour with a precision unavailable to earlier generations.

The danger is not that these systems will deliberately make everything boring, they just may become extraordinarily good at learning what causes objection.

If the easiest way to improve a metric is repeatedly to remove whatever divides an audience, optimisation acquires a direction. Language becomes safer, interfaces become familiar, products become broadly legible. Environments become easier to sell across markets.

None of those outcomes is necessarily bad in isolation, but what happens after several decades of them?

Markets are very effective at discovering what people will accept, but they are less reliable at preserving things whose value depends on being particular, locally specific or strongly preferred by only part of the population.

The disappearance of the colourful cars is almost trivial, the mechanism behind it is not. The choice is ours.