Substitute flavor: the market for fakes, simulations, and "almost-theres"

Copies, substitutes, imitations, and counterfeits have always been part of the market. In recent years, these formats have spread across very different categories and, at the same time, the boundaries between original, fake, "almost-there," and substitute have become more fluid, more contextual, and more dependent on familiarity and pattern of use.

The point of this piece is to look at all of this with less judgment (e.g., using a copy is tacky, changing a formula is dishonest, etc.) and to avoid simplistic explanations, ones that treat these phenomena as purely supply-side or purely demand-side. It's also, above all, about thinking through the kinds of relationships between brands and people that these strategies feed.

We have several paths to explore here: openly acknowledged substitutes, degraded originals, perceptible fakes, indistinguishable fakes, and things that are the same thing, but under a different name.

Openly acknowledged substitutes

The first type is the openly acknowledged substitute, which builds its value proposition through direct comparison with the original category. Perhaps the most literal example of this is the margarine I Can't Believe It's Not Butter, which references the category it imitates right in its name and implies a favorable comparison. Created in 1979 and still going strong today, despite changing hands multiple times (it was Unilever's from 1986 to 2018), it's arguably one of the most successful and longest-running examples of this positioning logic.

A more modern example, massively hyped a few years ago but now growing at far more modest rates than the 2020–2021 enthusiasm suggested, is meat substitutes. Within this category, perhaps the most emblematic product is the Impossible Burger, both for its technical effort at similarity and its emphasis on favorable comparison with the meat version: the word "impossible" directly alludes to the contrast between its ingredient list and the flavor it delivers.

The problem is that building a value proposition on top of the original category necessarily triggers the comparison, possibly an unfavorable one, immediately. To try to get around this, several NotCo products created an interesting twist on the pioneer margarine's strategy: negation (what psychologists would call establishing identity by contrast). The problem is that even negation still directly references the original category.

In Chile, this debate moved out of marketing and into the courts: NotCo had the sale of some products challenged, with an initial ruling in favor of restricting terms like "milk," later overturned on appeal. In Brazil, there was no direct legal confrontation, but the emergence and growth of these products was accompanied by a regulatory push requiring terms like "plant-based analog of," precisely to avoid confusion and create some distance from the reference to animal-derived products.

The core point: in many cases, the substitute isn't trying to deceive, in principle; it's trying to replace a known category that has some undesirable element while still claiming proximity to the desirable parts of it. It sells itself as a functional alternative, with claims that could be summarized as "X, but without the Y," as demonstrated here by Fabio (yes, that Fabio!):

• Margarine: the taste of butter, without sending your LDL through the roof

• Impossible Burger: the taste of a quality beef burger without the environmental impact / animal cruelty

• Contratypes (the honest ones): practically identical to that amazing fragrance you love, without costing a small fortune

If we use ecological relationships as an analogy for the type of relationship at play, it's mutualism between the customer and the brand producing the substitute; but it can be a form of "semantic parasitism" on the original categories, whose key attributes get appropriated. The judge of this dispute, as always, is the customer, who measures the distance between promise and delivery using the ruler of their own experience and available information.

Here's a reflection without a clear answer: all these examples evoke the original category or product. There's a heated debate in the marketing world about whether directly mentioning competitors is harmful, and that debate heated up considerably with Pepsi's Super Bowl ad this year, which showed the Coca-Cola polar bear doing a blind taste test. Does the same logic apply to categories? Does mentioning the original category clarify the concept, or does it doom the substitute to a game it starts out losing?

In our world of research and insights, the clearest fit here is synthetic personas (meaning the idea of qualitatively simulating real people using LLMs). With evidence mounting that the promise falls well short of reality, the constant need for updates with real data to keep them from becoming stuck in time, the absence of lived experience, the tendency to regress toward the mean and toward training data biases, they no longer look quite so promising. The industry's own discourse, which moved quickly from "user research without the users" to "might be useful for testing scripts and questionnaires" as scrutiny increased, suggests the semantic parasitism hit hard; even if they seem "real" to the uninitiated.

Originals in constant degradation

The second type is a near-original in steady decline, certainly the most common in consumer goods. It's the classic case of the product that keeps the category name and the brand while hollowing out its substance.

R&D team figuring out how much they can cheapen the formula without losing market share.

Driven both by cost and performance pressures and by the desire to shift tax classifications (examples: when certain categories receive more favorable tax treatment, the temptation is always to stretch the definition as far as it will go; see diary beverages and compounds, or the various fragrance families like EDP, EDT, and eau de cologne), consumer goods companies are constantly reformulating their products. In many cases, pushing the limits of what's acceptable to both consumers and regulators, moving toward what I'm calling "almost-theres": something already far from the original version, but not quite far enough to require a different name — a precarious place to be.

Two categories illustrate the possible market dynamics of these constant reformulations particularly well: beer and chocolate.

For a long time, the use of corn and other grains beyond barley has been part of the industrial logic of Brazilian beer. Because this wasn't clearly visible on the label (it appeared as "malted or unmalted cereals"), most people simply didn't know. In the 2010s, as premiumization was driven by craft beer and growing public interest, consumer knowledge started to rise. When a USP study came out showing the proportion of corn in popular beers, the issue went mainstream: it was scandalous because people didn't know, and they immediately associated it with low quality, which isn't necessarily true. Even so, the backlash was enough to force a reassessment of the category's messaging, with "100% malt" gaining a much more prominent position, even in mass-market products. In 2018, ingredients like corn became mandatory on labels.

Chocolate had already been on a trajectory of gradual ingredient cost-cutting, a trend dramatically accelerated by the cocoa price shock of 2023–2024. As a result, the criticism, previously more common in other countries and in foodie and elite circles, reached the Brazilian general public roughly last year. This broader public debate is directly connected to new regulation passed by Congress this April, which among other things raises the minimum cocoa solids content required to call something chocolate. It's not just here: a viral public complaint from the grandson of Reese's founder led the manufacturer to commit to returning to the original formula. Videos of "almost-chocolates" that turn rubbery and won't melt when heated have been circulating widely online, the result of compositions very different from what consumers expect.

The industry tends to frame this as mutualism ("we're managing to offer the product at a price consumers can afford"), but that's a narrative that only holds as long as the information asymmetry holds; in the eyes of anyone who sees what's actually happening, it can look like pure parasitism. These cases suggest the relationship needs that asymmetry to survive, given that the backlash came precisely when access to certain information arrived in both cases.

The research and insights version of this degradation shows up most clearly in online panels. After years of quality processes and controls that were flying by the seat of their pants, poorly audited externally and under heavy price competition, it eventually led to a major fraud scandal in the US. In some ways, this was a predictable problem: the raw material being squeezed here is the human being whose honest opinion we need, and they're the ones who have to come out ahead first. The quality and prevention discussion has become more visible in our market, especially there; but is it enough to address the lack of transparency and the participant experience?

The philosophical question of the "almost-theres," with no clear answer, is how much a category in a race to the bottom can be hollowed out before it has to be called something else, or collapses entirely.

Perceptible fakes

The perceptible fake is an imperfect copy that tries to pass as the original for at least some of the people who encounter it, often not the person who bought it. The obvious examples are imperfect copies of luxury goods and, more recently, electronics. Aesthetic prosthetics of all kinds also fit here: wigs, hair extensions, false lashes, dental veneers, and so on.

I'll use the ubiquitous knockoff JBL speakers blasting distorted audio, the kind you hear everywhere at street markets and beaches, as my example. This type of fake operates on the premise that, for someone who knows the original, the difference may be obvious, which makes its value highly dependent on each person's familiarity. At the same time, there's a group of people for whom they're "good enough," or for whom the quality gap doesn't justify the price gap. Another factor that often enters the equation is social judgment: whether there's a social cost to being seen or "exposed" as using something fake, and whether that cost matters to the user. This last characteristic is particularly interesting and unique to this type of fake when we're talking about physical products.

Because of this, the value architecture of this type of fake is among the most complex. A classical economist might apply a utilitarian justification: the perceived cost-benefit makes sense. An anthropologist or sociologist fond of Bourdieu might say that the value of things depends on the type of social signaling desired. A behavioral scientist might invoke cognitive biases to explain the gap between our interpretation and reality. All of these lenses are true, but none tells the whole story.

That said, and in defense of Bourdieu, as in that iconic scene from the first Devil Wears Prada, even when we say we don't care about something, we keep signaling and being read by what we signal; it's a system with no "outside" when we live in society. Still on the subject of social cost: we don't always see clearly what we're signaling. The classic comedy example is the guy wearing a wig who believes no one can tell. The intended signal doesn't match how it's perceived because it's entirely mediated by risk and familiarity.

Because of these particularities, the ecological relationships here are messier: fakes are clearly parasitizing the original products, and possibly also parasitizing the signaling of the person who buys them; but that effect depends on whether the fake is "exposable" in that context. And yet, in the buyer's mind, it's mutualism!

Three recent examples of perceptible fakes in research and insights:

• The study Claude called "the world's largest qualitative study", which, for anyone who knows the fundamentals of what qualitative actually is, is at best a quant study with open-ended questions in a conversational format.

• Social listening cosplaying as quantitative research: using open public mentions as if they were representative, and presenting data in percentages as if people had actually been asked. Embarrassing.

• A company publishes a study with a margin of error, run on their own proprietary panel. If the panel is proprietary, the margin of error tells you how well the study represents that panel, not the world at large. A textbook sleight of hand, and a brazen one at that.

Indistinguishable fakes

Unlike perceptible fakes, indistinguishable fakes hold up even to scrutiny from those with more familiarity, because they involve a far greater effort to reproduce the physical characteristics; the goal is to be as identical as possible, not just to seem similar.

There's a Chinese word for this concept that has caught on strongly there: píngti (平替), which roughly translates to affordable substitute. In English, people call them dupes or clones. The Brazilian press has been calling them superfakes. These are reproductions of much higher quality than ordinary fakes.

Anyone who thinks these high-quality copies are only made in China is very mistaken. In other countries where labor is cheaper, like Vietnam and Bangladesh, factory B-grades or copies produced on the same production lines and/or with the same materials as the originals are readily channeled into large regional marketplaces like Lazada or global ones like the well-known AliExpress.

They're invading a huge number of categories where manufacturing quality was an important barrier to entry and artificial scarcity is a major part of the price justification: handbags, watches, limited-edition sneakers, and countless others. One of the key barriers to perceptible fakes in these categories was precisely social judgment. With greater similarity, the framing becomes closer to that of openly acknowledged substitutes: the same physical appearance and the same social signaling, for less. Greater exposure on social media helps reduce the stigma and, in a way, normalizes it as a choice.

The rise of this type of fake brings an uncomfortable truth for anyone who believes branding solves everything: perhaps people aren't as willing to pay for a brand if the physical performance and the signaling capacity for others are equivalent, and the social risk is close to zero.

This logic of the perfect copy opens up new strategic paths:

• Copy as territorial defense: De Beers, one of the world's largest diamond miners, created in 2018 a separate synthetic diamond brand, Lightbox, selling at very low prices to deliberately commoditize lab-grown diamonds and signal to the market that they are something very different from natural ones, protecting the core business from cannibalization.

• Copy as political pressure. Culture jamming is the idea of using symbols, images, and communication codes in a critical and subversive way. That's what Mumumelon does, explicitly copying pieces from sportswear giant Lululemon, using alternative materials and processes, as a way of critiquing its environmental credentials, supply chain, and pricing.

• Separating the copy from the authentic as a service. In markets where copies are already very common, third-party authentication services exist as part of the value proposition, as with StockX or as an optional feature in targeted categories like eBay.

In research and insights, this seems to be the space that many players in synthetic data want to occupy, and our biases work in their favor, because they're convincing. The problem is that indistinguishable appearance is very different from indistinguishable quality; representing and understanding in the same way, especially for those who understand what lies beneath the visible surface, is another matter entirely.

The same thing, just reframed and renamed

This last type is perhaps the most insidious, because its practitioners take something that exists, replicate the content, but change the framing and/or the name; to escape the scrutiny or stigma present in the original versions, or even to sidestep regulatory constraints.

A recent example: testosterone replacement therapy (TRT), tied to legitimate medical needs, reframed as "hormonal optimization" when used for obviously aesthetic and sexual purposes. Unlike the standard clinical path, the patient journey and medical protocol arrive almost pre-packaged: the patient shows up at the doctor's office already wanting the thing, the doctor, known for practicing the thing, detects "evidence" that the thing is necessary and promptly prescribes it. It's a growing market in many parts of the world, including the United States, where distribution is already widespread, as well as in Brazil and the UK, where social media acts as a demand trigger. The latent demand is strong enough that hormonology markets itself as a reframing of endocrinology, to lend a veneer of legitimacy to the practice.

Another recent example: you're betting on the outcome of a military conflict involving human lives or on the use of nuclear weapons in the coming years; but relax, it's not gambling, it's prediction markets. There's a great deal of rhetorical gymnastics to explain the technical differences, but it's the new framing doing most of the work of separating one thing from the other; so much so that Brazil's CMN regulatory body vetoed them precisely because it didn't buy the narrative. No false moralizing here: betting isn't necessarily problematic. The issue is trying to polish the rawness of winning or losing money on events with enormous human costs, and calling it investing; without even getting into the minefield of information asymmetry.

This type of fake doesn't work by masking its characteristics like the previous ones; it pulls a slicker trick: it keeps almost all the signals in plain sight, but denies the identity, obvious to some, that would trigger harder scrutiny. It's like that Woody Woodpecker bit where he describes himself in perfect detail to the very person hunting him down, and then says he doesn't know the guy: "Yo no lo conozco, señor."

@previewcd

Não o conheço, senhor😄😄 #picapau #nostalgia #classico #desenhos

Research? No, señor, that's *cultural listening.*

The research and insights version of this happens when a process full of elements that clearly belong to the research universe, qualitative research in particular (active listening, interviewing or observing people, analyzing behavior vs. stated attitudes, etc.), gets rebranded with something glamorous enough to dodge the hard questions that the word "research" would attract: "sprint", "cultural listening", "intelligence blah-blah-blah," among many others. And then you go to the company's LinkedIn page and it's classified as market research, management consulting or information services. Sorry, what?

Final reflections

If we know and accept that reputation is built on repeated positive interactions, the hollowing out of substance in frequently purchased consumer goods seems like a particularly bad idea. Isn't that exactly how you alienate your most loyal customers, the ones most capable of noticing the difference?

Among all these alternative products, fakes, simulations, and reframings, part of the market manages to sell simulacra, almost-theres, and degraded versions; not because buyers consciously want less, but because they can't clearly see what's been taken away. There's nothing wrong with proposing alternatives; but if information asymmetry is what sustains the value proposition, the countdown to its implosion may have already begun: a problem for brands that should be thinking in decades, not quarters.

Thanks for reading to the end — see you in the next edition!