Liquid Death's Fans Left Over the Sweetener, Not the Marketing

Liquid Death is the brand every DTC deck reaches for. Canned water, a metal aesthetic, a sober-friendly tallboy at a festival where the only other option is beer, and a marketing operation good enough that people forward the ads voluntarily. If brand equity protects a company from anything, it should protect this one.

This is a voice-of-customer study of more than 8,000 real comments from people who drink, used to drink, or argue about Liquid Death. I went in expecting the usual brand-community pattern, where loud complaints sit on top of an intact base and the affection for the brand papers over the specifics. That is not what the comments look like. The complaints are specific, they name an ingredient, and a large share of them come attached to a competitor's name and a canceled subscription.

About this study

Four notes on scope before the findings, and they matter more than usual here because this is a single-brand study. First, a brand's own community overrepresents its most engaged fans and its most motivated critics, and underrepresents the far larger group who buy a can occasionally and have no opinion worth typing. Nothing in this corpus measures sales, market share or the size of the affected group, and it should not be read as doing so. Second, the sections vary a great deal in size. Sections 2, 4 and 5 rest on several hundred coded mentions each, while sections 1 and 6 rest on roughly 150 and 120, so read those two as directions rather than measurements. Third, this corpus is heavily weighted toward people who knew the earlier product. New buyers who only ever met the current formula have no comparison to make and are almost absent from the complaint conversation, which is a real limit on what the percentages here can tell you. Fourth, everything about ingredients, can sizes, discontinued flavors and company statements is reported here as what commenters said, not as verified fact about the company.

The research question

How have Liquid Death's sweetener change, can-size change and category expansion affected loyalty, and can provocative branding compensate when core fans believe the product itself got worse? The short version: the branding does not compensate, and the reason is more interesting than the failure. In this corpus the marketing was never load-bearing for purchase. It was load-bearing for identity, and identity turned out to be attached to a specific liquid in a specific can rather than to the company that sells it.

1. What the can was actually for

Start with what the product was doing for people, because it explains the shape of everything after it. Very little of this is about advertising. It is about a can being in the right place, doing a job that had no good alternative.

Second place is the emotionally important one and it is worth reading carefully, because it is not a brand preference. It is a social problem being solved by a physical object.

Notice what is doing the work in those four. A can that looks like a beer can, at a size that looks like a beer can, in a venue that sells beer. The brand's famous personality is the reason the object reads correctly across a crowded room, but the thing being bought is the object. First place makes the same point from the opposite direction.

2. What people say changed

This is the largest and least ambiguous ranking in the study. When people describe the product being different, they almost always name a single ingredient, and they name it with a precision that is unusual in consumer conversation.

The sweetener comments follow a repeated grammar: a statement of past affection, then the ingredient, then a verdict. That structure is what makes them so quotable and so damaging, because the affection is doing the work of establishing credibility before the complaint lands.

Third and fourth place are smaller but they compound, because they arrive at the same time as the first. A customer who has already decided the liquid is worse is not in a forgiving mood about paying the same money for less of it.

3. What people did about it

Complaining is cheap and this corpus has plenty of it. What makes the section useful is how much of the response is not complaint at all, but behavior described in the past tense with a named alternative attached.

The switching comments name competitors, which is the part a brand team should find hardest to read. Nobody is threatening to leave. They have already gone and they are telling other people where to go.

Second place has no equivalent in most churn studies and it deserves its own name. About one response in five is a customer performing quality control on the brand's behalf, learning to read the ingredient panel, tracking which retailers still hold older stock, and buying it when found.

The fifth position is real and should not be edited out. Roughly one coded response in eleven is someone who does not mind the change, and the most interesting of them is the person who has no memory of the earlier product to compare against.

4. Where the line extensions land

The sweetener change did not arrive alone. It arrived alongside an expansion into sodas, teas and caffeinated drinks, and in the comments the two are treated as the same decision. This ranking is a map of attention, not of approval.

The soda conversation is where the strategic argument actually happens, and the people making it are not hostile. Several are complimentary about the business logic while being certain about the cost.

There is a dissenting reading, and it is the one that makes the strategy defensible. If the goal is a bigger category, sweeter is correct, and at least one commenter says the change made the product better on its own terms while quietly conceding what it cost.

That last comment is the whole section in two sentences. The product may well be better. It is also no longer the product the category-defining customer was buying, and the sparkling water conversation elsewhere in this corpus supports that: people choose a seltzer specifically because it is not sweet. The same tension runs through our study of why people use electrolyte products outside exercise, where a category built on a functional claim keeps drifting toward tasting like a soft drink and losing the reason people picked it.

5. The marketing everyone admires, and the product they left

Now the question the whole study exists to answer. The marketing conversation about this brand is large, sophisticated and largely positive about the craft. It is also almost entirely disconnected from whether the person typing intends to buy anything.

Read the admiration comments closely and the pattern is consistent. The praise is offered as an assessment of the company's skill, in the third person, often immediately before or after a sentence explaining why the commenter does not buy it.

One commenter in the marketing community offers the sharpest version of the argument, and it is not really about advertising at all.

The most quoted exchange in the corpus is an argument about brand loyalty itself, prompted by what commenters describe as the company's response to the complaints. Both sides of it are worth reading together, because between them they explain why the marketing could not do the work here.

6. What still holds the ones who stayed

The last section is the smallest, roughly 120 coded mentions, so treat it as a direction. It is also the most practical, because it is a list of the things that survived.

The plain unflavored can is doing an enormous amount of work here. It is the one product in the range that commenters agree did not change, and a meaningful share of the remaining loyalty has retreated onto it.

The format comments are the ones with an expiry date attached, and they are the clearest illustration of how a loyalty anchor becomes a churn trigger when it is removed.

The question this corpus could not answer

Whether any of it shows up in the numbers. A brand community is a gathering of the people who care most in both directions, and nothing in 8,000 comments tells you how many customers were affected, how much volume moved, or whether the larger category the company went after more than covers the segment it upset. It is entirely possible that this was a good trade and that the loudest version of the loss is confined to a small, articulate group who were never going to be the growth story. What the corpus does establish is the shape of the risk, and the shape is worth knowing on its own: the exits were product-level, specific, and completely unmoved by a brand that most of the people leaving still openly admire.

A customer glossary

If you work on brand, product or lifecycle for a consumer brand with a passionate community, these are the terms already in circulation among the people you sell to.

Frequently asked questions

Did Liquid Death's marketing protect it from the reformulation backlash?

Not in the language people use. Of about 114 coded comments where someone describes changing their buying behavior toward the brand, roughly 66% name something about the product itself, meaning the sweetener, the taste, the can size or the price. Only about 8% name anything about the marketing, the advertising or the company's public conduct. The admiration for the brand's marketing is real and appears in the same threads, often in the same comment as the exit. It just does not function as a reason to keep buying.

What do Liquid Death customers say actually changed about the product?

Overwhelmingly the sweetener. Of coded mentions describing a change, about 57% name the move away from agave and toward stevia, about 21% describe the taste and the aftertaste it leaves without naming the ingredient, about 10% describe a flavor that was pulled or is no longer findable, about 9% describe the price or the can getting smaller, and about 3% describe the carbonation or something else about the liquid. Stevia is named by name in roughly 750 comments across the corpus, close to one comment in ten.

What did customers do after the Liquid Death formula changed?

They moved, and they moved specifically. Of coded responses to the change, about 44% describe switching to a different sparkling water brand, usually a named one such as Spindrift, Waterloo, Polar or Topo Chico. About 21% describe hunting for remaining old-formula cans and reading ingredient panels before buying. About 16% describe stopping altogether, several by canceling a standing Amazon subscription. About 9% wrote in to the company, and about 9% say they stayed and do not mind the change.

How do drinkers describe Liquid Death's move into sodas, teas and energy drinks?

As a category change rather than a range extension. Of coded mentions about the product line, about 39% discuss the flavored sparkling waters, about 31% the sodas, about 16% the iced teas, about 9% the energy and caffeinated line, and about 4% the plain water. The recurring argument in the soda conversation is not that the sodas are bad. It is that the brand left a category it owned for one it does not, described by one commenter as leaving a market segment they dominated for one where they are a tiny fish in a big sea of soda.

What do people actually think of Liquid Death's marketing?

The marketing is respected and the price is resented, at the same time. Of coded mentions in the marketing and brand conversation, about 66% are some version of paying a premium for water, about 13% discuss the sponsorships and the celebrity turn, about 10% state plainly that the marketing is genuinely good, about 8% say the edge feels forced or has worn off, and about 3% say the branding is the actual product. Notably, the admiration is usually offered as a compliment to the company and not as a reason the commenter buys.

What still keeps loyal Liquid Death customers buying?

Product attributes, not brand attributes. Of coded reasons people give for still buying, about 35% point out that the plain unflavored water was never part of the change, about 25% describe the large cold can as a format nothing else matches, about 18% describe the job it does at a bar or a show when they are not drinking, about 16% cite how widely it is stocked now, and about 5% say they simply want the brand to be good again. Every one of the top four is something a competitor can copy with a production decision.

This report analyzes consumer language and perceptions about a named beverage brand. Statements about ingredients, formulations, can sizes, discontinued products, pricing, retail availability and company communications are reported as claims made by commenters at the time of writing and were not independently verified with the company. Nothing here is a statement of fact about the composition, labelling or safety of any product, and nothing here is health, dietary or medical advice. Descriptions of physical reactions to any ingredient are individual consumer accounts, not evidence of a general effect. Brand and product names appear only as they were mentioned by commenters and remain the property of their owners. Percentages are directional estimates from this corpus, not a census of customer behaviour, and this study measures conversation, not sales.