Amazon rating system developers ended up being too protective of it

According to the former Amazon executive, the online ratings and reviews model was a good solution at first, but the team ended up being too protective of it.

Dan Lewis spent many years at Amazon before founding his own company. Among the biggest projects he worked on was helping to build an online ratings and reviews model.

It was a big problem from the online retail giant in its early days.

“The original problem we were trying to solve at Amazon was customer confidence. You can’t touch, see, smell — how do you get confidence to buy?” Lewis explained at the recent CNBC Technology Executive Council summit in New York City.

Peer reviews, with a rating of one to five stars, solved the original customer confidence problem. But in that success, Lewis says he learned a key lesson about setting oneself up for failure. In a nutshell: falling in love with solutions rather than never-ending problems.

“Oftentimes you protect the thing you built versus remembering the existential problem you are trying to solve”.

“We didn’t name the group that works on the problem the ‘customer confidence team.’ We named it the ratings and reviews team,” Lewis said.

Amazon employees like Lewis who worked on the original solution became convinced their job was to make ratings and reviews better, instead of thinking of other solutions, and newer technologies, that could solve the problem better.

More Amazon news