How do you decide what to charge for your products or services? This is one of the most frequent questions I hear at conferences and inside corporate offices. It’s both a question and a howl of pain, because buried inside it are so many unspoken fears: Can I escape the race to the bottom? How can I increase my margins? How do I stay profitable?

Pricing is typically discussed as though it were some kind of black magic, an art you can’t possibly understand unless you specialized in marketing while doing your MBA. Then you hire one of those graduates only to discover they don’t know either. But two business school professors at the University of California at San Diego have an answer: Experiment!

Uri Gneezy is one of the most imaginative business school professors you’ll find. He’s done a whole range of weird and wonderful experiments, all of which have shown him that a well-designed trial is faster and more effective than a month of speculation and strategic guesswork. When I met him a few weeks ago in San Diego, he was full of excitement about a recent assignment—to figure out how to price the wines made by a local producer.

They weren’t wine experts but, together with his wife—assistant professor of marketing Ayelet Gneezy—the two academics designed an experiment: on different days, they priced the Cabernet Sauvignon sold at the winery at three different points—$10, $20, $40. Apart from the price change, everything else about the customers’ experience of the winery remained the same: the selection, the brochures, the progression from white to red to dessert wines. Only the price varied. At what price point would the vineyard sell the most Cabernet?

It turned out to be: $20. Selling the wine cheaply didn’t make people buy it more but value it less. The experiment – which had cost the winemaker virtually nothing – increased his profits by 11 percent.

Gneezy is a gifted designer of experiments; some of his forays into human competitiveness and pricing are among the most thought-provoking in behavioral economics. But his main argument is more important than any of them. Don’t think you know – or have to know – everything. Experiment. Gather data. Base your decisions on empirical evidence.

I’ve always resisted the idea that business is a science. When Jim Collins talks about his ‘lab’ my skepticism comes out in a rash, because you can’t do controlled experiments comparing companies. But what Uri and Ayelet Gneezy have demonstrated is that you can do controlled experiments within your own business. You can also do pilots: small forays into new ways of working which you can measure and learn from. In my own experience, pilots taught us more, faster than any amount of discussion and debate.

Using pilots and experiments yields two valuable and inspiring outcomes. First, it means that you can present decisions based on data. That means everyone implementing those decisions can understand where they came from. But the second outcome is at least as important: it relieves the CEO of the burden of being omniscient. You don’t have to know everything. You can experiment and find out.

Experiments beat market research because they’re real, they’re fast, and if you’re disciplined enough to test just one thing at a time, they will yield data you can act on. Big corporations could do them too—and sometimes do. The problem, the Gneezys found, was that the data often gets lost in bureaucracy and politics. Entrepreneurial, private businesses shouldn’t have this problem. So the prayer mat and dartboard can be retired today.

 ‘Life as a Lab’ by Uri Gneezy and John A. List will be published in the summer.