Tetra

Tetra invented flake fish food in the 1950s and, in doing so, invented the home aquarium as something ordinary people could keep alive. The company has not been in German hands since 1974, and has been sold four times over. The name on the tub never changed.

In 1954, in a laboratory in Lower Saxony, a biologist registered a trademark: a water flea [21].

If you kept tropical fish in 1954, the water flea was your problem. Daphnia — a translucent crustacean a millimetre or two long — was what tropical fish ate. You netted them from ponds, or bought them from someone who did. They were seasonal, they were awkward, and they carried disease into the tank along with the protein. Keeping exotic fish alive in a German living room meant running a supply chain of live crustaceans, and it was the largest single reason the hobby stayed small [11].

The biologist was Ulrich Baensch, and he had spent the previous year in a laboratory he had named BioMin, working on the thing that would make daphnia unnecessary [22]. He got it: a food that could be manufactured, boxed, shelved and sold. By 1955 TetraMin — the first dried flake food for ornamental fish — was on sale worldwide [11].

And then he put the water flea on the label.

It is still there. In Tetra’s own words: „1954 haben wir den Wasserfloh als Warenzeichen für unsere Marke eintragen lassen. Seitdem fungiert er als unser Gütesiegel“ — in 1954 we registered the water flea as a trademark for our brand; since then it has served as our seal of quality [21]. Seventy years on, the mark of quality on a tub of Tetra food is a picture of the animal that tub exists to render obsolete.

Baensch went further. When he moved the company to the small town of Melle in 1962 and launched a magazine for the hobby, he called it the Wasserflohpost — the Water Flea Post [10][11].

There is a particular kind of confidence in naming your customer newsletter after the thing you killed.

The thesis

Tetra is usually filed as a company that quietly dominates an obscure market. That undersells it. Tetra did not find the ornamental fish food market and win it — there was no market to win. It manufactured one. The only peer-reviewed review of ornamental fish nutrition credits “the development of manufactured feed since 50 years ago” with “the tremendous growth of this hobby” [2]. Hermann Simon, the man who coined the term hidden champion, put Tetra’s share of that world market at over 50% in 1996 [1].

Thirty years later it is still the name on the tub. But it has not been German-owned since 1974, it has been sold four times since, and it now sits inside an American conglomerate that does not disclose it as a line of business at all. Tetra is a hidden champion with nearly every trait of one stripped away — except the winning.

The Niche

A tub of fish flakes looks like the least demanding product in the pet shop. It is closer to the opposite.

Start with the animal. “Ornamental fish” is not a species; it is a merchandising category. Combine feeding habit (herbivore, omnivore, carnivore) with temperature regime, salinity and water hardness, and you get “more than 18 distinguishable nutritional groups of ornamental fish” [2]. Their requirements are not close to each other. A juvenile goldfish needs around 29% crude protein; discus need 44.9–50.1% [2]. Life stage moves the target as violently as species does — goldfish larvae need 53% against 29% for juveniles, a 24-point swing inside a single animal [2].

Now put them in the same box. A community tank holds surface feeders, mid-water feeders and bottom feeders at once, with different daily rhythms; the food has to satisfy all of them from one shake of one tub [2]. Particle size is a trap in both directions — small enough for the little fish to swallow, large enough for the big ones to register as food [2].

The academics who reviewed the field reached a conclusion that reads exactly like TetraMin’s design brief, written forty-eight years after the product: “A better approach than to satisfy the needs of individual species would be to minimise deviations, thus to develop an intermediate feed that would satisfy the crude requirements of all species” [2].

Then there is the part that makes it genuinely hard: the food is eaten underwater, and it begins dissolving on contact.

This is called leaching, and the numbers are brutal. Up to 50–70% of a feed’s residual vitamin C activity can be lost after ten seconds of immersion, along with 5–20% of its pantothenic acid and up to 27% of its folic acid [5]. For commercial flake diets specifically, a large percentage of the vitamin C, vitamin B12, choline and pantothenic acid is gone within thirty seconds of feeding [2]. And the aggravating factor is an indictment of the entire product format: “the smaller the feed particle size and the longer the feed remains uneaten in the water, the greater the loss of water soluble vitamins through leaching” [5]. A flake is close to the worst possible geometry — maximum surface area, minimum mass.

Vitamin C is the sharpest case, because the fish cannot help. Teleost fish have lost gulonolactone oxidase, the enzyme catalysing the final step of vitamin C synthesis, so every molecule they get must arrive in the food [7]. It is also the nutrient that dies fastest in manufacturing: processing and storage losses for unprotected L-ascorbic acid in practical fish feeds run “as high as 95%” [5]. And the failure is invisible — juvenile Oscars fed an unsupplemented diet take twenty-five weeks to show deformed jaws and haemorrhaging [2], which is roughly six months after the owner would have stopped connecting it to the food.

So the brief is: one product, for eighteen nutritional groups at once, delivered as a thin floating particle that starts shedding its most fragile and most essential vitamin within ten seconds of hitting the water, where the deficiency it causes takes half a year to become visible and the customer will blame something else.

The Drum

The way Baensch solved it — and the way almost all flake food is still made — is drum drying. A slurry of ground ingredients is fed onto a heated rotating roller, the water flashes off, and scraper blades peel a dried sheet off the drum surface, which is then broken into flakes.

We know what was wrong with that process in unusual detail, because Tetra wrote it down. Its own patent for a replacement reads like a confession of forty years of constraint [3]:

• The roller runs at a steam temperature of 120 to 165 °C.
• The flake strip comes out “about 0,15 mm” thick and “cannot be varied very much”.
• Final moisture “can only be adjusted within narrow limits”.
• And therefore: “heat-labile raw materials, for example enzymes and probiotics, cannot be used.”
• The flakes come out uneven, in differing sizes, needing sieving.

A second patent names the limit that cost the most money. Fat-rich mixtures “stick on the roller driers normally used and cannot be removed to give a comminutable foil strip” [4]. The consequence was that ornamental fish food ran at 0.5–8% fat while aquaculture feeds reached up to 25% [4]. For decades, pet fish were underfed on energy because of the behaviour of grease on a hot cylinder.

Tetra’s escape was to stop using the drum. Extrude a formed body at under 70 °C with a water content of 40% or less, then roll it flat [3]. Everything the drum had fixed becomes adjustable again: thickness can be set anywhere from 10 µm to 5 mm, moisture from 1 to 30% [3].

And that unlocks something neater than it first appears. Thin flakes float longer; thick flakes sink faster. So by blending thicknesses in a single tub, one product can feed surface, mid-water and bottom feeders simultaneously [3] — the community-tank problem solved not by chemistry but by a manufacturing tolerance. Three products’ worth of work done by a roller gap.

The vitamin C answer is printed on the label, if you know to look. American tubs of TetraMin list L-Ascorbyl-2-Polyphosphate — a phosphate ester of vitamin C, engineered to survive the heat and the water that destroy the plain molecule [5].

One detail on that patent is worth pausing on. The assignee is Warner-Lambert [3]. By the time Tetra reinvented the manufacture of the product it had itself invented, the company was a subsidiary of an American pharmaceutical group, and the patent went in under the parent’s name.

Which brings us to the part of this story that is not about fish.

The Origin

Ulrich Baensch got his first aquarium at nine and his doctorate in natural sciences at twenty-five, in Kiel. He began breeding fish and aquatic plants in 1947, set up a wholesale business called Tropenhaus in 1950, and opened a retail shop in Hannover in 1951 — the registration that counts as Tetra’s founding [9]. The BioMin laboratory followed in 1953 [22], the water-flea trademark in 1954 [21], TetraMin in 1955 [11].

The name has nothing to do with the neon tetra, or with any fish. Tetra is Greek for four — the four flake types in the original formula — bolted to the last syllable of Vitamin [11]. The most-fed aquarium fish in the world eat a food named after a number.

Here is what I could not find, having looked hard in two languages: a single direct quotation from Ulrich Baensch. The man invented a product category and left no voice in the public record at all.

What he left instead was the infrastructure of a hobby. The Wasserflohpost became the Aquaristik-Fachmagazin and is still published today; more than a hundred books in twelve languages came out under the Tetra imprint [10]. Baensch sold the food and also supplied the literature that taught people how to want it. His son Hans founded a second publishing house, Mergus, in Melle in 1977, and produced the Aquarien Atlas, whose first volume ran to 992 pages [20] — for a generation of German fishkeepers, the reference work.

The company moved to Melle in 1962 with about twenty employees, renamed Tetra Kraft Werke — “Tetra Power Works”, which sounds like a municipal utility. By then Baensch had filed fifteen manufacturing-process patents and twenty trademarks [22][9]. By 1975 there were around 400 employees on 16,000 m² of production floor [11].

There are 389 today [11]. Fifty years, and the payroll has not moved.

The Sale

In 1974, twenty-three years after founding it, the Baensch family sold Tetra to Warner-Lambert, the American pharmaceutical company [11]. The price has never been made public, and I could not find it. It is a strange gap: the exact moment this hidden champion stopped being one is the least documented event in its history.

What follows is a chain of owners with no connection to fish:

• 2000 — Pfizer acquires Warner-Lambert in a roughly $90bn merger, and inherits a fish food company as a non-core asset.
• December 2002 — Pfizer sells it to Triton Partners, a European private equity fund, in a deal reported at around $240m [13].
• April 2005 — Rayovac, the battery company, buys Tetra Holding GmbH for approximately €415m [12]. It renames itself Spectrum Brands weeks later.

Pfizer sold it at the end of 2002 for around $240m; twenty-nine months later Rayovac paid approximately €415 million for it [12][13]. In any currency, private equity comfortably more than doubled its money. At the point of that sale Tetra had approximately 700 employees, generated about €200 million in annual net sales, and operated in over 90 countries [12]. The buyer knew exactly what it was purchasing. “The Tetra brand is arguably the most recognized global brand name in the pet supplies industry,” said United Pet Group’s president at the time [12].

The family kept the books, not the food. Mergus Verlag was founded three years after the sale. The Baensch name survives in German aquarium publishing; it does not survive on the tub.

The Market

Here is the number that makes this company strange.

In 2025, every brand of ornamental fish food sold through German bricks-and-mortar retail, added together, came to €60 million. Cat food came to €2,315m and dog food to €1,753m. Fish food is the smallest of the five pet food categories the trade association measures — about 1.4% of the German stationary pet food market [16].

The world leader of this category leads the smallest aisle in the shop.

It is also an aisle that has not grown in a very long time. The German fish food market was worth €60m in 2010 [8]. It was worth €60m in 2025 [16]. Fifteen years flat in nominal terms is a real-terms decline of roughly a third.

The installed base underneath it is eroding, but unevenly, and the unevenness is the whole business model. German garden ponds stocked with ornamental fish fell from 1.3 million in 2023 to 1.0 million in 2025 — down nearly a quarter in two years. Aquariums went from 2.2 million to 2.1 million. Meanwhile spending on fish hardware — tanks, filters, pumps, lighting — fell about 10% across the same three years, while spending on fish food was flat [16].

People are not setting up new tanks. They are still feeding the ones they have.

That is a good business to be in, and the economics of the product make it better. TetraMin flakes weigh about 200 g to the litre [23]. A 250 ml tub — 52 g of food — retails at €4.19, and the 10-litre catering bucket at €39.49 [19]. Do the division and the same product ranges from roughly €19 to €80 a kilogram depending on how much of it you buy at once. The small tub, which is what almost everyone actually buys, sits at the top of that range: fish flakes at the price of a decent jamón.

Nobody notices, and the reason is in the nutrition literature. A neon tetra at maintenance eats about four milligrams of feed a day [2].

Those two facts come from different places and the join between them is mine, not anyone else’s — but it is the whole business model. At four milligrams a day, a four-euro tub is years of feeding for a small fish. An €80/kg product is invisible when the unit of consumption is a milligram, and a price nobody perceives is a price nobody shops.

The Moat

Run Tetra against Hermann Simon’s traits and it passes most of them comfortably. Narrow focus, deep rather than broad: yes. Globalisation from a tiny home base: a single plant in small-town Lower Saxony supplies the brand worldwide — Spectrum Brands’ own annual report confirms that “most Tetra® aquatics nutrition and care products [are] manufactured in our Melle, Germany facility” [14], and at the 2005 sale Tetra was already operating in more than 90 countries [12]. Innovation intensity: the Melle site runs roughly 700 test aquariums holding some 35,000 live fish, and every new feed is observed for twelve weeks before it goes near a shop [8]. Closeness to the customer: the company published the hobby’s magazine and its manuals.

Two traits it fails outright, and they are the two Simon cares most about: independence and leadership continuity. Tetra has been family-owned for 23 of its 75 years. Its managing directors on the German commercial register are the parent group’s officers — the chief financial officer, the general counsel — not fish people in Lower Saxony.

So why does the position hold?

Not, I think, because of the flake. The flake is good, but it is a compromise and always was. A 2021 feeding trial put TetraMin head to head with a rival in tanks holding two ornamental species together and found opposite optima in the same water: the neon tetras did best on TetraMin alone, while the rasboras did better on an alternating diet. The authors’ conclusion is that feeding co-housed, anatomically diverse ornamental species is “a highly ambiguous task” [6]. The universal flake is still exactly what the academics described — an intermediate feed that minimises deviations. It is simply the best-executed one, for so long that the compromise is now what the market means by “fish food”.

The moat is really three things, and only one of them is technical.

The category is too small to be worth attacking. Sixty million euros a year in Germany does not justify anyone else’s programme of species-specific nutrition research across eighteen nutritional groups. Spectrum’s own filing declines to claim any market share for Tetra and describes pet supplies as “highly fragmented with no competitor holding a substantial market share” [14] — though note that is the whole pet supplies category, not fish food, and the two should not be blended.

The purchase is unexamined. See the milligram arithmetic above. Price competition requires a customer who notices the price.

And the brand taught the category. When you inherit a hobby whose founding literature was published by the company whose food you buy, “which brand” is not a question that comes up.

Spectrum’s chief executive stated the commercial consequence more plainly than any analyst could, on an earnings call in November 2025: “We’re the world’s largest player in Tetra. We have the best brand. You know? It’s recognized without having to advertise” [15].

A brand that no longer needs advertising, in a category too small for anyone to contest. That is the whole moat in one sentence, said out loud by the owner.

The Cracks

The most candid statement of the risk also came from that call, from the same man: “Kids like to live on these iPhones all day long. You know, they don’t like taking care of fish tanks” [15].

The striking thing is that Tetra’s own European managing director had said the same thing fourteen years earlier, in a German regional paper, in four words: „Aquaristik muss cooler werden“ — fishkeeping has to get cooler [8]. The company has understood its demographic problem for well over a decade and has not solved it.

The rest of the risks are in the parent’s filings and they are specific:

Customer concentration. Amazon and Walmart each exceed 10% of Global Pet Care segment sales and together account for approximately 34% of it [14]. Two customers, a third of the segment.

Private label. Spectrum names it explicitly as a cause of the segment’s fiscal 2025 decline, alongside general category softness [14]. A commoditising category with an unexamined purchase is precisely where a retailer’s own brand does well.

Tariffs, pointed straight at Melle. The filing states that the company “manufactures aquatics nutrition products at its facility in EMEA and imports such products into the U.S., which are also subject to the enacted tariffs” [14]. The single German plant that supplies the world is now a tariff exposure.

The recall. In May 2025 a single lot of Tetra ReptoMin was voluntarily recalled after internal testing found possible Salmonella in one component of the blend; no illnesses were reported [18]. Set that against what Tetra’s managing director told a newspaper in 2011: „Hier hat es nie einen Futtermittelskandal gegeben“ — there has never been a feed scandal here [8].

And the quiet one: invisibility. Spectrum Brands discloses no aquatics revenue line anywhere in its annual report — the Global Pet Care segment is broken out by geography and revenue type, and no further [14]. Nobody outside the company can size this business. In 2018 Tetra closed its Melle logistics centre and moved it to the Netherlands, costing around 60 jobs in the town [17]. A business that cannot be sized from outside is a business whose decisions arrive as news.

Takeaways

Inventing the complement beats winning the category. Baensch did not compete for share of fish food; he removed the reason the hobby stayed small, then sold into the hobby he had enlarged. The most durable position is upstream of the market you end up dominating.

A brand can outlive every other component of a hidden champion. Family ownership, independence, continuity of leadership, even national ownership — Tetra has lost all of them and kept the position. Whatever the moat was, it was not the founding family, and the hidden-champion literature does not entirely want that to be true.

Price invisibility is an underrated moat. €80 a kilo is unshoppable when the customer buys it four milligrams at a time. If your product is consumed in quantities too small to notice, you are competing on habit, not on price — and habit is much harder to attack.

When the installed base shrinks, be the consumable. German fish hardware fell 10% in three years while fish food stayed flat. The tank is a discretionary purchase; the food is not. In a declining category, the recurring item is the safe half of the business.

---

One for the replies: Tetra’s trademark is a picture of the water flea it put out of business — a company advertising itself with the thing it made obsolete. What other product is quietly named after, or drawn from, the thing it replaced? I can think of a few. Hit reply; the best ones go in the next issue.