A startup mailed millions a cat-shaped scanner to swipe magazine ads open to websites — nobody wanted to scan a cat to do what a URL already did, and the CueCat became the dot-com's punchline.
At the height of the dot-com boom, a company called Digital Convergence raised roughly $185 million and built a small plastic gadget shaped like a cat. Its job was to solve a problem: when you saw an advertisement in a magazine, on a billboard, or on television, how did you get from that ad to the company's website? Digital Convergence's answer was the :CueCat — a barcode scanner you plugged into your PC, so you could swipe a special barcode (a "cue") printed beside the ad and have your computer open the related page for you. To get it into people's hands, they gave it away free, by the million: Forbes mailed 830,000 of them to subscribers, Wired sent out more than 500,000, and RadioShack handed them out in stores. Over ten million were manufactured. It is now routinely cited as one of the worst gadgets ever made — and it is one of the cleanest design lessons in the history of consumer technology.
The problem it solved was already solved
Start with the question the CueCat was built to answer: how do I get from a printed ad to a website? In 2000, the answer was already sitting at the bottom of the ad. You read the URL — "visit example.com" — and you typed it. It took two seconds, required no hardware, no installation, no registration, and no company in the loop. The CueCat proposed to replace that two-second habit with a physical device tethered to your computer by a cable, a software install, and an account.
This is the original sin, and it is a design sin, not a technology one. The device did, with friction and a gadget, exactly what people were already doing effortlessly with their fingers. Walter Mossberg said it plainly in the Wall Street Journal in October 2000: "In order to scan in codes from magazines and newspapers, you have to be reading them in front of your PC. That's unnatural and ridiculous... the device fails miserably." A product that adds steps to an action people perform without thinking is not an innovation. It is a tax on a behavior that was already free.
You had to come to the desk
The deepest flaw was geometric. Magazines, newspapers, billboards, and television are things you encounter away from your computer — on the couch, on the train, in the kitchen, on the road. The CueCat required you to be wired to a PC at the exact moment you wanted to scan. So either you carried your magazine to your desk and read it next to a tethered cat-shaped scanner, or you did what everyone actually did: you ignored it.
The setup tax made it worse. Before you could scan a single cue, you installed the :CRQ software, connected the device as a keyboard "wedge" through a PS/2 port, and registered — name, age, email, ZIP code, gender, plus a survey about your shopping habits and education. Reports put the full setup at about an hour. An hour of effort, and a permanent fixture on your desk, to do something a URL did in two seconds from anywhere. The friction wasn't a rough edge on a good idea. The friction was the idea.
And it quietly watched what you scanned
Then there was the cost nobody mentioned at the cash register, because it wasn't paid in cash. Every time you scanned a cue, the device sent the product code, your user ID, and the scanner's serial number back to Digital Convergence. The "free" gadget was a tracking instrument: the business model was the data trail of what millions of people pointed it at. In September 2000 that bargain blew open — security researchers found that the company had left roughly 140,000 registered users' personal details (names, email addresses, age range, gender, ZIP code) sitting unencrypted on a world-readable web page. The company's response was to offer affected users a $10 RadioShack gift certificate. The hacker community, meanwhile, gutted the thing: people published drivers that stripped out the registration and let the scanner read any ordinary barcode, defeating the surveillance the whole venture was built on.
So the full proposition, stated honestly, was: install our software, tether our gadget to your PC, register your personal details, and then do — with more steps — what typing a URL already did, while we record everything you scan. Phrased that way, the only mystery is how $185 million flowed toward it.
The money, and the collapse
A great deal did. Belo Corporation put in $37.5 million, RadioShack $30 million, Young & Rubicam $28 million, Coca-Cola $10 million, with General Electric and E.W. Scripps among the rest — $185 million in total behind a unit that cost about $6.50 to make. The arithmetic of giving ten million of them away for free only works if the data and the click-through traffic are worth a fortune. They were not, because almost nobody used the device. In May 2001 Digital Convergence fired most of its 225-person staff; that September Belo wrote off its entire $37.5 million stake and the cues stopped appearing in print. By the end of 2001 the original system was effectively dead. Gizmodo later voted the CueCat the single worst invention of the 2000s; Time put it on its "50 Worst Inventions" list; PC World ranked it among the worst tech products of all time. A cat-shaped monument to a problem that didn't exist.
The design-intelligence reading
The CueCat is the textbook case of a product that adds hardware, steps, and hidden cost to do what people already do effortlessly — and effortlessness is not a small advantage you can engineer around. It is the whole game. The competitor the CueCat had to beat was not another gadget. It was a two-second habit, performed from anywhere, for free, that nobody had to be taught. Against an incumbent behavior that cheap and that frictionless, a tethered scanner with an hour of setup and a registration form never had a chance — and the cleverness of the device (it really did decode the cue and open the page) was completely beside the point. The engineering worked; the judgment failed.
This is exactly the judgment design intelligence exists to make before the hardware is tooled and ten million units are shipped. The discipline is to start from how people actually behave — where they are, what they already do, how much friction they will genuinely tolerate — and to be honest about the thing you are really competing against, which is almost always an existing habit, not a rival feature list. Adding a step, a device, or a hidden tracking cost to a task that is already effortless is not innovation; it is a tax, and people refuse to pay it no matter how slick the gadget. The cleverest peripheral on earth still loses to a behavior that takes two seconds and costs nothing. Respect the habit, or watch your gadget die on the desk beside the keyboard — a little plastic cat, gathering dust, pointed at nothing.
Sources
Related posts
Harley-Davidson put its rebel badge on perfume and cake-decorating kits — and nearly diluted the toughest brand in America into a punchline.
Microsoft Bob replaced the desktop with a cartoon house and a "helpful" dog that explained things you already knew — users hated being patronized, and it was gone in a year (but it left us Clippy and Comic Sans).
Pepsi spent a fortune to make a clear cola signal "purity" — then shoppers' brains refused a cola that didn't look like one, and Crystal Pepsi vanished in under two years