Posted on by omarmetwally

What Lego and Magna-Tiles Teach Us About Who Actually Owns an Idea

By Oskar Rausch

First Published: 18 July 2026

Written from a ranch in the hills of Mendocino County, where the sun comes close enough to touch and 140 acres of solitude echo the sounds of success

There is a particular kind of object that fools you.

It looks simple. A colored block. A flat plastic tile. You look at it and think: a child could have designed this. And that thought — that specific, seductive thought — is the trap. Because the simplest-looking objects tend to have the most complicated histories, and the gap between who invented a thing, who owns a thing, and who got rich from a thing is where fortunes are made, lost, litigated, and quietly stolen.

I want to tell you about two toys.

Both are in your house right now, or were, or will be. Both are considered iconic examples of their country of origin. Both are wrong about that. And both of them ended in a courtroom teaching the same lesson from opposite directions.

Neither story ends where you think it does.

PART ONE: THE DANISH BRICK THAT WASN’T DANISH

I. The Fairy Tale

Everyone knows the Lego origin story, because Lego tells it very well.

A Danish carpenter named Ole Kirk Christiansen, in the town of Billund, starts making wooden toys during the Depression. He names the company from the Danish leg godt — “play well.” His son takes over. The family perseveres. The bricks conquer the world. Four generations later it’s still family-owned, the largest toy company on earth, and a genuine cultural institution.

Every word of that is true.

It is also missing the part where the brick came from somewhere else.

II. Hilary Fisher Page

The man you’ve never heard of is named Hilary Fisher Page, and he deserves better than he got.

Born in Sanderstead, England on 20 August 1904, the first child of a father who worked in the lumber trade, Page made his own wooden toys as a child and was educated at Shrewsbury School from 1918 to 1923. He was, by every account, a serious man about a serious subject. A biographer records that he spent the whole of every Wednesday in a different nursery school, sitting on the floor and playing with the children, to find out exactly what kind of toys would interest them most.

That detail matters. Page wasn’t a manufacturer looking for a product. He was a child-development obsessive looking for a better toy. He had become increasingly unhappy using wood as a material for children’s toys and was an early advocate of plastics as a safe and hygienic alternative. In 1936 he began manufacturing Kiddicraft “Sensible” toys using new injection moulding technology. This is what motivated him to make toys from plastic rather than wood — they were safer and cleaner for children to play with.

Then he invented the thing.

In April 1939, Page applied for British patent #529,580, “Improvements in Toy Building Blocks,” which was accepted in November 1940. It described hollow square bricks with four studs on top. The bricks were hollow rectangular plastic bricks with four round studs on top; the sides of the blocks could fit between the studs of another block, allowing children to build up structures such as walls.

Read that description again. That is a Lego brick. In 1939. In England. Patented.

Page kept going. He patented the basic 2 × 4 studded brick design in 1947, followed by patents for the side slits in 1949 and the baseplate in 1952 — designs now featured in exhibits at the Brighton Toy and Model Museum. The Kiddicraft Self-Locking Building Brick sets were first marketed in 1947, and as a promotion Page and his family built large display models for the 1947 Earl’s Court Toy Fair. His daughter Jill remembered building “Sky Scrapers” that were as tall as she was — five foot six.

There is a photograph in my head of that scene that I cannot shake: a man and his daughter, at a toy fair in postwar London, building a plastic skyscraper taller than she is, out of a brick he invented, for a company that will be a footnote within a decade.

III. The Copy

Here is how it happened, and the mundanity of it is the point.

Kirk Christiansen, the founder of Lego, came across these bricks in a demonstration he was shown of an injection moulding machine. He then copied the bricks and sold them under his own brand as “Automatic Binding Bricks.”

Not corporate espionage. Not a stolen blueprint. A sales demo for a plastics machine. The machine vendor needed something to mould to show what the machine could do, and what they had lying around was a Kiddicraft brick.

Ole and Godtfred would copy both the Interlocking Building Cube (as Lego Plastic Byggeklodser) and the Self-Locking Building Brick (as the Automatic Binding Brick).

Was it illegal? Here the story gets genuinely murky, and I want to be fair to the Danes. One argument holds that Lego technically stole nothing, since in Scandinavian countries the Self-Locking Building Bricks were never patented — a patent is a national instrument, and a British patent does not restrain a Danish manufacturer. It has also been noted that the Self-Locking Building Bricks were just one of a bunch of products created by Page, and one of the less successful ones, with poor sales — British kids preferred the more versatile Minibrix.

So: a commercially struggling product, unprotected in Denmark, copied by a man who saw more in it than its inventor’s market did.

That’s not theft. That’s something more uncomfortable. That’s the ordinary operation of the patent system, working exactly as designed, producing a result that feels wrong.

And this is the part that should make every inventor’s stomach turn over:

Page died without ever finding out that Lego had copied his product.

He died on 24 June 1957, aged 52. He never knew. The most successful toy in human history was being manufactured in Denmark from his design, and the man who spent his Wednesdays sitting on nursery floors went to his grave without the information.

IV. What Lego Actually Invented

I don’t want to be unfair to Godtfred Kirk Christiansen, because he did invent something, and it was the thing that mattered.

Page’s bricks had a flaw. The problem Page faced with early plastics production was tolerance and grip. In order to grip well — a property now called “clutch power” — the space a brick’s stud fitted into had to be deliberately fractionally too small, so the walls of the receiving brick had to flex to accommodate the stud. That flexing meant the receiving brick’s walls were effectively “sprung” to grip the stud. If the space is a perfect fit for the stud, or fractionally larger, the brick doesn’t hold.

Page’s bricks didn’t hold well enough. Kids’ towers fell down.

After Lego adopted the same style of bricks as Kiddicraft, they made many improvements, such as adding tubes inside the bricks to make them stick together. That is the stud-and-tube coupling: the hollow tubes on the underside that grip the studs from within. It solved the tolerance problem, and it is the reason a brick manufactured in 1958 still clicks satisfyingly into one manufactured this morning.

Godtfred filed for it in January 1958.

So the honest accounting is this: Page invented the brick. Lego invented the grip. One of those men is a footnote and the other founded a dynasty, and the difference between them is not the quality of the idea.

V. The Purchase

Some 31 years later, Lego acquired Kiddicraft — while they were preparing to, ironically enough, sue Tyco for illegally copying their bricks.

Sit with the choreography of that. Lego is about to go to court to stop a competitor from copying its brick. Its lawyers, doing diligence, look at the chain of title on the brick itself. And someone in a conference room in Billund says: we should probably own Kiddicraft before we walk into that courtroom.

They bought the ghost so it couldn’t testify.

PART TWO: WHAT HAPPENED IN COURT

This is the part of the Lego story nobody tells, and it’s the part that actually matters if you own a patent.

I. Interlego v. Tyco (1988)

Tyco Industries and its Hong Kong subsidiary entered the model-building market in 1983, producing bricks compatible with Lego by reverse engineering Lego products. Lego sued Tyco in Hong Kong for copyright infringement of the design drawings used to manufacture Lego and Duplo bricks.

Note the theory carefully, because it’s clever and it’s desperate. Lego did not sue over the brick. Lego sued alleging infringement of copyright in its engineering drawings — not the bricks themselves.

Why? Because the patents were gone. Lego’s rights were originally based on designs by Hilary Page, whose patents expired in the 1950s; Lego’s own improvements, including the third-generation brick with enhanced clutch-power, were protected by patents and registered designs until 1975.

Patents expire. Copyright, in 1988, lasted the author’s life plus fifty years. So if you could argue that your engineering drawings were protected artistic works, and that anyone making a compatible brick had indirectly copied those drawings — you could resurrect a dead monopoly and give it another century of life.

Lego also argued that its post-1972 drawings, which largely copied the earlier drawings with minor technical amendments, qualified as original artistic works entitled to a fresh term of copyright.

Redraw the drawing, restart the clock. Forever.

On 5 May 1988, the Judicial Committee of the Privy Council — Lord Keith of Kinkel, Lord Templeman, Lord Ackner, Lord Oliver of Aylmerton and Lord Jauncey of Tullichettle — decided the case.

The Privy Council dismissed Lego’s appeal and allowed Tyco’s cross-appeal, with the result that Lego’s claim of artistic copyright infringement was dismissed except with regard to two minor items. No copyright subsisted in Lego’s pre-1973 drawings, and the post-1972 drawings did not qualify as original artistic works entitled to copyright. Consequently, Tyco’s manufacture of compatible bricks did not infringe Lego’s copyright except in limited cases. Lego’s attempt to extend its monopoly beyond patent and design protections through copyright claims on minor drawing modifications was rejected.

The court’s reasoning is worth quoting in substance: after Lego’s patent and design protection expired, it claimed copyright protection in drawings of the blocks which included a number of minor variations from the original block design, and the Privy Council dismissed the claim that the revised drawings constituted an artistic work.

Lego lost. Tyco was allowed to continue selling the bricks, which at the time were earning them about $20 million annually.

II. Kirkbi v. Ritvik (2005)

Seventeen years later, in Canada, Lego tried the other door.

Kirkbi AG owned patents for Lego toy building blocks, the last of which expired in Canada in 1988. After the patents expired, Mega Bloks began marketing a line of blocks identical in size to Lego blocks, using the same geometrical pattern of studs on top coupled with tubes underneath.

The plaintiffs commenced a passing-off action under s. 7(b) of the Trade-marks Act, asserting unregistered trademark rights with respect to the “Lego indicia” — the distinctive geometrical pattern of raised studs on the upper surface of the block.

The theory: even if we can’t patent the studs anymore, the studs are so recognizably ours that they function as a brand. A trademark. And trademarks, unlike patents, never expire as long as you keep using them.

The Registrar of Trade-marks refused registration, so Kirkbi claimed the Lego indicia as an unregistered mark and relied on s. 7(b) and the common law doctrine of passing off.

On 17 November 2005, the Supreme Court of Canada, in unanimous reasons by LeBel J, dismissed the appeal and held that Kirkbi’s passing-off claim was barred by the doctrine of functionality.

The Court held that Kirkbi’s design was purely functional, and that “a purely functional design may not be the basis of a trademark, registered or unregistered.” The Lego indicia trademark was primarily functional and therefore invalid; the plaintiffs were attempting to extend their expired patent protection through the guise of a trademark, which was contrary to the policy underlying trademark law.

The ruling had a profound effect on the toy industry, enabling competitors to produce compatible building-block products without fear of trademark infringement claims based on functional shapes, and directly facilitating the market entry and expansion of rivals like Mega Bloks in Canada. It also reverberated internationally, informing Lego’s ongoing litigation against Mega Bloks in the European Union and the United States, where courts after 2005 drew on Kirkbi’s logic.

Two courts, two continents, seventeen years apart, same answer:

You get twenty years. You do not get to launder an expired patent into a permanent right by calling it something else.

III. So Why Is Lego Still Winning?

Here is the fact that makes the whole thing click.

Lego lost the copyright case in 1988. Lost the trademark case in 2005. Its patents have been expired for half a century. Anyone on earth may legally manufacture a brick that snaps onto yours.

Lego had revenue of 83.5 billion Danish kroner — about $12.6 billion — in 2025. It is owned 75% by Kirkbi and 25% by the Lego Foundation, and employed 31,282 people as of 2024. Kjeld Kirk Kristiansen and his three children have a combined net worth of nearly $25 billion. Kjeld’s own net worth was approximately $10.2 billion as of November 2025.

The monopoly is gone. The money is bigger than ever.

Because of what the family built while the patent was still alive: Kjeld oversaw the expansion of Legoland theme parks in the 1990s and began the practice of licensing branding rights from moviemakers, which led to the release of Star Wars-themed building sets in 1999. The company also expanded into computer games, films and clothing.

That’s the answer. Mega Bloks can copy the stud. Mega Bloks cannot copy Star Wars. Mega Bloks cannot copy Legoland. Mega Bloks cannot copy the fact that a brick from 1958 fits a brick from today, a compatibility promise maintained across seven decades of manufacturing discipline that no new entrant can retroactively acquire.

The patent bought twenty years of cover. What Lego built under that cover is the actual business — and it turns out that what it built was not protected by the patent at all.

PART THREE: THE JAPANESE TILE WITH THE GREEK NAME

Now the mirror image.

I. Pythagoras

If you have a small child, you have stepped on a Magna-Tile barefoot in the dark and briefly reconsidered your entire life. They’re the translucent magnetic polygons that snap edge-to-edge into glowing little cathedrals. They feel American in a specific way — the earnest, STEM-forward, wooden-Montessori-adjacent educational toy.

They are Japanese.

Magna-Tiles were originally developed in Japan, where they were sold under the name Pythagoras. Magna-Tiles is the U.S. version of an educational toy called Pythagoras, which People Corporation began selling in Japan in 1992. The trademark “Pythagoras” (ピタゴラス) was applied for in Japan by People Corporation Ltd. on 25 December 1991 and registered on 31 May 1995, Japanese registration number 2707566.

II. The Salesman

American salesman Rudy M. Valenta saw the toy while visiting Japan in 1996 and bought the rights to it. He founded Valtech Co. in 1997 with his wife Noriko, to bring the magnetic tile sets to the US and the world.

And then the single most instructive decision in this entire essay:

At first the product was launched in the US under the name Pythagoras, but it had little traction, as the branding and name did not match the US market. It was quickly rebranded to Magna-Tiles, and carried the Pythagoras logo on the packaging for a few years before it was dropped permanently.

Read that as an IP practitioner and it’s almost too neat. The product was identical. Same tiles, same magnets, same geometry, same everything. The only variable changed was the name, and the name was the difference between a product with little traction and this:

Sales began slowly, but by 2015 Magna-Tiles were being described as “ubiquitous” among young children in the United States, and the New York Times and other publications have highlighted them as a top toy.

Rudy Valenta did not invent the tile. He did not invent the magnet. He did not invent the geometry, the edge-capture mechanism, or the educational theory. He invented a word, and the word was worth more than the invention.

III. What Valtech Actually Enforces

Now watch what they defend in court — because it is not the tile.

In litigation in the Northern District of Illinois, Valtech alleged that defendants were selling transparent, colorful plastic building blocks in various sizes and colors almost identical to Magna-Tiles under the brand names “Magnetic Stick N’ Stack” and “Playmags.” Rudolph Valenta, owner of Valtech, holds a federal registration for the mark “Magna-Tiles” for use in association with toys, blocks and puzzles — U.S. Reg. No. 2,654,320. Valtech also claimed that the packaging used by the defendant was identical, or at least confusingly similar, to the packaging used for Magna-Tiles.

The court’s own summary of the claims describes them as trade-dress violations, with Valtech competing for sales against products that infringe on Magna-Tiles’ trade dress.

Trademark. Trade dress. Packaging. Not the tile. The name and the look.

This is not an accident or a weak litigation strategy. It is the only ground they have, and they know it, because the underlying concept was never theirs to lock up. The magnetic-tile category is now a crowded field — similar products include Magformers, Playmags, CONNETIX and Picasso Tiles — and the patents in the space are held by scattered parties. To pick one at random: U.S. Patent 10,857,475, issued 8 December 2020, covers a magnetic-tile game set with holding strips to prevent buckling, plus accessories like figurines and lighting — a genuine invention, owned by someone who is not Magna-Tiles.

And yet:

Magna-Tiles bills itself as America’s #1 Magnetic Building Sets Brand, operating out of Bolingbrook, Illinois, and in March 2025 revealed its first fan-made invention at the American International Toy Fair.

The MAGNA-TILES trademark is now held by MVW Holdings, Inc., filed 22 April 2018, covering toy building blocks, toy construction blocks, and construction toys incorporating magnets.

No foundational patent. Number one anyway.

PART FOUR: THE LESSONS, STATED PLAINLY

I’ve been on both sides of this. I have been the man with the clever mechanism and the pending application, convinced the mechanism was the thing. And I have watched, more times than I care to count, the money flow past the cleverest mechanism to whoever controlled the system, the name, and the factory.

Here is what these two toys, laid side by side, actually teach.

1. A patent is a candle, not a sun.

Twenty years, and then it is gone, and your competitors are legally entitled to everything you disclosed. That is not a bug — the disclosure is the deal. You told the public how it works; in exchange you got two decades of exclusivity. When the term ends, the public collects.

The Privy Council said no in 1988. The Supreme Court of Canada said no in 2005. Lego, one of the best-resourced litigants in the toy industry, tried to extend an expired monopoly through copyright and then through trademark, and lost both times, because courts are specifically alert to that maneuver.

Plan for the expiry from the day you file. If your business dies when the patent does, you don’t have a business. You have a countdown.

2. The winner is whoever owns the commercial rights and the story — not the inventor.

Hilary Fisher Page invented the interlocking brick and died not knowing it had conquered the world. A Japanese math-toy company invented the magnetic tile and is invisible in the American market it created. The Christiansens and the Valentas won.

This is not a morality tale about theft. Lego improved the brick meaningfully. Valenta took real risk importing an unproven product. But neither of them originated the core idea, and both of them captured essentially all of the value, and if you are an inventor you need to internalize why: they controlled distribution, brand, and manufacturing. Those are not lesser assets than invention. In the long run they are the only assets.

3. In the “deceptively simple” category, provenance is the whole game.

Both of these products live in the most dangerous neighborhood in IP: ideas that look too obvious to own. Construction systems. Snapping parts. Elementary geometry.

That is precisely the category where an earlier, obscure, foreign design surfaces at the worst possible moment. Lego’s origin has a permanent asterisk on it for exactly this reason — even the Privy Council’s own recitation of the facts begins by noting that Lego’s rights were originally based on designs by Hilary Page. Forty years after the fact, the ghost was still in the record.

If you are filing in a simple-mechanism space, prior-art search is not a formality you delegate and forget. It is the load-bearing wall.

4. Build a “tell.”

Lego has clutch power. Magna-Tiles has its lattice and its name.

The tell does double duty. It is functional, and it is a signature — an instantly legible marker that says this is the real one. Note carefully, though, what Kirkbi teaches about the limits: a purely functional design may not be the basis of a trademark, registered or unregistered. A tell that is nothing but function will not survive as trade dress. It needs an arbitrary, aesthetic, non-necessary component — something you chose when you didn’t have to.

That non-necessary flourish, the part an engineer would call waste, is the part that can outlive your patent by a century.

5. Name it before you perfect it.

The Pythagoras-to-Magna-Tiles rebrand is the cheapest, highest-return act in either story. Zero engineering. Zero tooling. A word.

Inventors systematically underweight this because naming feels like marketing, and marketing feels like the thing you do after the real work. It isn’t. It is the asset with the longest half-life you will ever create. The tile is copyable. The mould is reverse-engineerable. The patent expires on schedule.

The name is forever, if you feed it.

CODA

The candle burns for twenty years. It is bright, and it is genuinely useful, and it makes you feel like you own something permanent.

You don’t. You own time. That’s all a patent has ever been: a fixed quantity of protected time, purchased with public disclosure.

What you do inside that window — the brand you build, the factory tolerances you dial in, the promise you make to customers and then keep for seventy years, the word you put on the box — that’s the part nobody can take when the candle goes out.

Hilary Fisher Page had the better claim and died anonymous. Rudy Valenta had no claim at all and built an empire on a rebrand.

The difference wasn’t the invention.

Out here on the ranch, where the silence is complete enough to hear your own thinking, that’s the thought that keeps circling back. The simplest brick. The hardest lesson.

SOURCES

Kiddicraft & Hilary Fisher Page – Wikipedia, “Hilary Page” — biography, dates, Kiddicraft founding, plastics advocacy – hilarypagetoys.com, “K281 Kiddicraft Interlocking Building Cubes” — British patent #529,580, applied April 1939, accepted November 1940 – Brickipedia, “Kiddicraft” — 1940 patent, brick geometry, Lego’s tube improvement – Shameless But Famous Wiki, “Kiddicraft Blocks” — 1947 basic-brick patent, 1949 side slits, 1952 baseplate, Earl’s Court Toy Fair – Brickfetish.com, “1947 — Kiddicraft” — Page’s nursery-school research, Jill Page’s recollection, Lego’s copying of both product lines – Brighton Toy and Model Museum — clutch-power tolerance problem, Page patent images (GB587206, GB633055, GB673857) – atlbrick.com — Cellulose Acetate composition, 1945 slot improvement, plastics rationale – TodayIFoundOut.com — injection-moulding demo origin, Scandinavian non-registration, Kiddicraft acquisition, Page’s death without knowledge

Interlego A.G. v. Tyco Industries Inc. [1989] AC 217; [1988] UKPC 3 – Wikipedia, “Interlego AG v Tyco Industries Inc” — 5 May 1988, panel composition, full citations – CaseMine — full holding, Page patent lineage, Tyco reverse engineering from 1983, pre/post-1973 drawings – vLex Canada — appeal dismissed, cross-appeal allowed, artistic-work analysis – CaseMine commentary — Seventh Schedule analysis, functional vs. aesthetic boundary – lawprof.co — Paragraph 8(2), Schedule 7, Copyright Act 1956

Kirkbi AG v. Ritvik Holdings Inc., 2005 SCC 65, [2005] 3 SCR 302 – Wikipedia, “Kirkbi AG v Ritvik Holdings Inc” — 17 November 2005, LeBel J unanimous, doctrine of functionality – Supreme Court of Canada, decisions.scc-csc.ca — official reported decision – vLex Canada — Canadian patent expiry 1988, Lego indicia claim, extension-of-monopoly reasoning – Smart & Biggar, “LEGO Loses Trademark Battle in Canada” — the “purely functional design” holding – DWW, “A Leading IP Case: Lego® and Mega Bloks®” — constitutional and functionality holdings – Grokipedia — downstream effect on Mega Bloks market entry and international litigation

Magna-Tiles / Pythagoras / Valtech – Wikipedia, “Magna-Tiles” — Japanese origin, Valenta 1996 discovery, Valtech 1997 founding, rebrand, 2015 ubiquity, competitor list – Hattori & Partners (Japanese patent attorneys), “MAGNA-TILES” — People Corporation 1992 sales, Pythagoras trademark JP Reg. 2707566, filed 25 Dec 1991, registered 31 May 1995 – Valtech, LLC v. 18th Ave. Toys Ltd., 14 C 134 (N.D. Ill. Feb. 12, 2015) — trade dress claims, U.S. Reg. No. 2,654,320, packaging allegations – govinfo.gov, official court PDF of the above – Justia Trademarks — MAGNA-TILES held by MVW Holdings, Inc., filed 22 April 2018 – U.S. Patent 10,857,475 (Justia) — third-party magnetic-tile patent, issued 8 Dec 2020 – ZoomInfo / PRNewswire releases — Bolingbrook IL, “America’s #1” positioning, March 2025 Toy Fair

Lego Group financials & ownership – Wikipedia, “The Lego Group” — 2024 revenue 74.3bn DKK, ownership split, headcount – Wikipedia, “Kirkbi” — holding structure, subsidiaries, 2023 generational transfer – Bloomberg Billionaires Index (Kjeld Kirk Kristiansen; Sofie Kirk Kristiansen) — 2025 revenue 83.5bn DKK / $12.6bn, 75% Kirkbi stake – Forbes profile, Kjeld Kirk Kristiansen — ownership split, succession – Fortune / Yahoo Finance — combined family net worth ~$25bn, Kirkbi net income – Grokipedia — $10.2bn net worth as of November 2025

A NOTE FROM THE AUTHOR

I want to be careful about the moral framing here, because it would be easy to write this as a heist story and it isn’t one.

Ole Kirk Christiansen did not steal Hilary Page’s brick in any sense a court would recognize. He saw an unprotected foreign product at a machine demo, recognized something in it that its own inventor’s market had not, and then spent years fixing the engineering flaw that had kept it from working properly. That is what innovation frequently looks like from the inside — not a lightning bolt, but somebody noticing an underperforming thing and being unreasonably stubborn about improving it.

Rudy Valenta didn’t steal anything either. He bought the rights, legitimately, and took the risk of importing a toy that had already flopped once under its original name.

The discomfort in both stories isn’t villainy. It’s the system working correctly and producing an outcome that feels unjust anyway. Page’s protection didn’t extend to Denmark, because patents are national. The Japanese inventors sold their commercial rights, because that’s what rights are for. Nobody broke a rule.

And yet one man died not knowing, and another man’s fortune is measured in the billions.

If there’s a reason to keep telling these stories, it’s this: inventors are systematically miscalibrated about which asset matters. We overvalue the mechanism and undervalue everything downstream of it — the name, the factory, the promise, the decades of boring consistency. We treat the patent as the prize when it is merely the starting gun.

I write this from a place where you can see a long way in every direction, which turns out to be a useful condition for thinking about time horizons. Twenty years is not very long. Ask anyone who’s watched a patent expire.

— Oskar Rausch, Mendocino County, California

Nothing in this essay is legal advice. If you have a pending application in a crowded mechanical art, retain a competent patent attorney and pay them properly. The one thing every story in this piece has in common is that the cost of good counsel is trivial compared to the cost of finding out late.