In 1999, the start-up Webvan launched in the US with the intention of becoming the grocery service for online shoppers. It climbed to a multibillion dollar valuation at its IPO in late 1999, and declared bankruptcy less than two years later.

This disaster is all the more remarkable because Webvan's founder, Louis Borders, was an experienced entrepreneur who had built up a successful chain of book shops with his brother. He hired heavyweight management, poaching the chief executive of Andersen Consulting (now Accenture), and secured backing from Goldman Sachs and leading venture capital firms such as Sequoia and Benchmark.

But Webvan had spent massive funds in a "get big fast" strategy intended to seize a dominant piece of the market. Without having thoroughly tested its business, Webvan invested $1 billion on distribution centres. Even as Webvan was losing massive sums of money on its all-or-nothing gamble, the UK grocery giant Tesco had been cautiously testing and developing its own business delivering groceries over the internet. In 2001, the same year that Webvan folded, Tesco announced online sales of £146 million. According to Tesco's 2014 annual report, 86 per cent of its product sales are now made online.

Data and experimentation

There were plenty of other bad bets by impetuous entrepreneurs and their investors between the mid-1990s and early 2000s, with the result that the Nasdaq’s value crashed in October 2002 to just a quarter of its peak in March 2000.

Yet herein lies a silver lining. As a result of the dotcom bust, there has been a change in outlook among tech entrepreneurs. They began to move from dogma to rationalism: where once an entrepreneur might pursue an idea on blind faith, the untested, unproven hunch was no longer adequate. Passion began to matter far less than proof.

Tech entrepreneurs have flocked to the “lean start-up” approach. They use customer relationship management systems to understand their users’ behaviour from a distance; they build sales funnels that track when customers are ready to buy and what influences them, and they carefully monitor how customers react to subtle tweaks to their services with the objective of constantly improving them. In short, they are treating business not as a vocation or a vision, but as an experiment to be tinkered with and optimised.

They are essentially doing what Feargal Quinn was doing in his shops decades ago. The Superquinn founder quickly recognised the importance of repeat custom in the grocery business. Instead of maximising profit per visit, which was the norm, he emphasised the importance of drawing customers back repeatedly.

Quinn designed his shops so that managers had nowhere to hide. There was no office where they could take phone calls or do paperwork. They could be accosted by a customer on the floor at any time. Quinn himself spent Wednesday mornings with panels of customers at Superquinn, hearing what they thought of the shop. In other words, Quinn’s managers faced the customers directly every day and, on Wednesdays, he faced them himself in the hope of finding things he could improve.

Hubris versus the prototype

In contrast, consider a far more humble entrepreneurial example. Several years ago, I co-founded a technology venture that focused on tour guide operators. The product we had conceived would radically change how tourists perceived new cities, and would also bring new marketing benefits to tour operators. We were convinced of the idea’s greatness, which is why we kept it quiet and told nobody about it.

We drew together a team to build a demonstration of the product, and kept what we were doing quiet. During a long and difficult development process, we never thought of interviewing tourists or tour operators. Our appetite for criticism was nil. So rather than build something basic and test it with users, we strove for a version that would have the maximum impact.

The venture came to naught, primarily because we were trying to build something to too high a specification, but more importantly because we had not taken the time to ask if what we were building would be relevant to the people we hoped to sell it to.

For all its current woes, Tesco’s trial-and-error approach trumped Webvan’s all-or-nothing gamble for the same reason that Feargal Quinn’s shops engaged repeat customers: incremental experimentation works. The first step towards the world’s best idea is to be willing to abandon the world’s best idea.

Johnny Ryan is executive director of the UCD Innovation Academy