Among the current crop of ex-Microsofties building software startups, Emmanuel Marot is unique. Not just because he hails from Niort, a town in western France, by way of Paris (or for his impeccable fashion sense), but because he is trying to do something fundamentally new with Internet commerce.
It’s called scientific pricing, and it works something like this. The price of commodities is at the core of commerce. But prices are still typically set by trial and error—by guessing, and marking up by some amount over the cost to the seller. When you buy an espresso (Marot had two during our meeting last week), who’s to say whether you should pay $1.75 or $1.85? It turns out the right pricing can benefit both buyers and sellers.
This by itself is not new, of course. In the late 1970s, airlines started using a technique called revenue management to jigger their prices and boost sales, while also helping consumers get better deals. Hotels, car rentals, utilities, and retailers have followed suit. What’s new about Marot’s work is that he’s trying to bring this sort of automatic, “dynamic” pricing to the masses—for anything you might buy or sell in a classified ad on Craigslist, for instance. His new startup, Valu Valu, is rolling out its public beta Web-based software the week of February 23.
Marot and his business partner, Bruno Botvinik, moved to Seattle in September 2006 after Microsoft bought their Paris-based mobile search company, MotionBridge. Marot and Botvinik left Microsoft in May 2008 (“shortly after they removed our golden cuffs,” says Marot) to start working on Valu Valu. Why get out of mobile search? “It’s hard to make a serious business in mobile,” Marot says, in part because the market is driven by big, slow-moving carriers.
For the past couple of years, Marot has been reading about auction theory and the principles that underlie pricing. The need for something better than eBay or Craigslist became real in the fall of 2006 when he was trying to sell his son’s bike online. Should he ask for $80? $70? Marot got no offers until it dropped below a certain price ($60), but why was that? And the same thing for more expensive items, like a used kayak he eventually sold for $1,900. If he could come up with an algorithm to determine
