# BackRub Dies, Google Lives
In 1996, two Stanford PhD students built a search engine that crawled the web by analysing backlinks — the invisible threads connecting one page to another. They called it BackRub. The name was accurate. It described exactly what the technology did. And it was, by any measure of branding science, catastrophically wrong.
BackRub lasted about a year before Larry Page and Sergey Brin replaced it with a misspelling. They wanted googol — the mathematical term for 10 to the 100th power, a one followed by a hundred zeros. Sean Anderson, a fellow grad student, typed "google.com" into a domain registrar to check availability. The typo stuck. Within seven years, the misspelled word was processing over a billion searches per day. Today, Alphabet's market capitalisation hovers above $2 trillion, and the word "google" has been a verb in the Oxford English Dictionary since 2006.
The question worth asking isn't why Google succeeded. It's why BackRub had to die first.
## The Phonetic Autopsy of BackRub
David Placek, founder of Lexicon Branding and the linguist behind names like BlackBerry, Pentium, and Swiffer, has spent four decades studying why certain names create commercial velocity and others don't. His framework analyses names across dimensions that most founders never consider: opening letter energy, vowel flow, processing fluency, and what he calls the compound multiplier.
BackRub fails on nearly every dimension. The B opening carries reliability — BMW, Boeing, BlackBerry all benefit from that plosive solidity — but it's immediately undercut by the soft, retreating "ack" vowel. The compound structure (back + rub) triggers associations that are, at best, irrelevant to search technology and, at worst, uncomfortably physical. The name creates cognitive friction: the listener's brain has to work to suppress the massage association before it can process the intended meaning.
Processing fluency research, published across journals including the Journal of Consumer Research and replicated in studies at Stanford, consistently shows that names requiring additional cognitive processing are remembered less accurately and evaluated less favourably. BackRub doesn't just fail to describe search — it actively competes with its own product for the listener's attention.
## The Accidental Genius of a Typo
Google, by contrast, is what linguists call a trochaic form — a STRONG-weak rhythmic pattern. GOO-gle. The stress falls on the first syllable, which is the dominant pattern in English and the one that human brains process with the least effort. Apple. Facebook. Twitter. Amazon. The trochaic bias in English means that names following this pattern feel natural before anyone consciously analyses them.
The double-O is rare in technology naming. In 1998, tech names clustered around hard consonants and Latinate suffixes: Compaq, Novell, Netscape, Infoseek. The OO sound — a back vowel that perception studies consistently rate as larger, rounder, and more playful — gave Google a phonetic identity that was impossible to confuse with anything else in the category. Rory Sutherland, vice chairman of Ogilvy and author of Alchemy, calls this the "costly signal" theory of naming: a name that sounds unusual signals confidence, because only a company sure of its product would risk sounding different.
The G opening carries what Placek's research categorises as hard consonant confidence — a voiced plosive that creates a sense of authority. The -le ending, meanwhile, is a diminutive suffix in English, the same pattern that makes "little," "cuddle," and "giggle" feel approachable. Google is simultaneously authoritative and friendly. It's a grown-up in a playground. That tension is not accidental — though in this case, it was not designed either. The typo produced what a professional naming team might have taken months to engineer.
## The $23 Billion Test
When Google went public in August 2004, the company achieved a market capitalisation of $23.1 billion on day one. The IPO was unconventional — a Dutch auction that bypassed Wall Street's traditional allocation system — but the name had been doing unconventional work for six years before that.
Consider the competitive landscape Google entered. In 1998, the dominant search engines were AltaVista, Lycos, Excite, Infoseek, and Yahoo. Each of these names either described the product (search, seek, find) or tried to sound futuristic through Latinate construction. They were naming within the category. Google named outside of it.
Scott Galloway, professor of marketing at NYU Stern and author of The Four, argues that the most valuable brands in technology share a common trait: they create their own semantic territory rather than competing within existing territory. Apple doesn't mean fruit anymore. Amazon doesn't mean river. Google doesn't mean number. When a brand name colonises a word completely, displacing its original meaning, that's what Placek calls cumulative advantage — and it's nearly impossible to reverse.
The companies that competed with Google on descriptive naming are gone. Infoseek was absorbed by Disney. AltaVista was shuttered by Yahoo. Excite went bankrupt. Lycos was sold for $36 million after peaking at a $5.4 billion market cap. The search engine that won wasn't the one with the clearest name. It was the one with the most distinctive one.
## The BackRub Trap
The instinct to name descriptively is powerful, and it's almost always wrong for companies that intend to grow beyond their first product. BackRub described a technology. Google described nothing — and that's precisely why it could become everything.
This is Sutherland's central insight about naming: logical names satisfy the namer, not the market. When a founder names their company after what it does today, they're optimising for internal clarity. When they name it something distinctive and slightly uncomfortable, they're optimising for external recognition. These two goals almost never converge.
The research supports this. A 2003 study by Yorkston and Menon at Stanford demonstrated that brand names with sounds matching product attributes (a "fast" sounding name for a fast car) performed better than descriptive names in recall tests. But Google goes further — it doesn't match its product's attributes phonetically. It creates entirely new associations. The name doesn't tell you what Google does. It tells you how Google feels: expansive, playful, limitless.
## What a Typo Teaches
Every founder who names their company after its core feature is making the same bet that Page and Brin initially made with BackRub: that clarity beats distinctiveness. The evidence from two decades and $2 trillion in market capitalisation suggests otherwise.
The lesson isn't that you should misspell a word and hope for the best. The lesson is that the names which feel wrong in a conference room — the ones that don't describe the product, don't reference the technology, don't play it safe — are often the names that create the cognitive space a brand needs to grow.
BackRub was a perfectly accurate name for a perfectly accurate technology. Google was a perfectly inaccurate name for a company that needed room to become something no one could yet imagine. The misspelling wasn't the genius. The willingness to abandon accuracy was.