retailciooutlook

Amazon's Secret Weapon in the Fight to Improve Product Information

By Stephanie Leffler, CEO, OneSpace

Stephanie Leffler, CEO, OneSpace

Amazon has grown their ecommerce empire by building infrastructure to power their operation and then releasing that infrastructure to other companies in the form of services in the cloud. Known as Amazon Web Services, this division of Amazon is responsible for $2.6 billion dollars of their Q1 2016 revenue representing 64 percent year-over-year growth.

"Amazon realized they needed human intelligence, but believed there was a more scalable, more cost effective alternative to a large full-time or outsourced workforce"

While Amazon is best known for providing services that impact a company’s hardware and hosting infrastructure, they have followed this model in other areas as well. One lesser-known but equally interesting offering is known as Amazon Mechanical Turk.

Like the rest of the services under the AWS umbrella, Mechanical Turk was built to solve one of their own large scale problems. Amazon was working to find new ways to clean up and enhance product information. Products needed to be properly described, categorized, tagged with attributes and de-duplicated.

Software could help, but algorithmically attacking this problem across thousands of different data sources and web pages was practically impossible. Amazon realized they needed human intelligence, but believed there was a more scalable, more cost effective alternative to a large full-time or outsourced workforce. They resolved to leverage the growing wave of freelancers looking for online work opportunities and they began to build a system that would allow them to distribute work to thousands of people in parallel, making their catalog enhancement process better faster and cheaper than any of their competitors. This led to the development of Mechanical Turk as we know it today—and to the rapid improvement of Amazon’s product data.

As 2016’s total ecommerce spend is expected to approach $22 trillion dollars, consumers have come to expect more from online retailers. A robust selection of products is a must, causing many retailers to adopt marketplace models to add scale their assortment. At the same time, consumers expect better, more complete product information. Eighty-eight percent of consumers say that “product content is extremely or very important to their purchase decisions,” according to a recent study by Salsify.

While consumers are clear in their need for better product content, retailers have not invested enough in augmenting and enhancing product information to meet consumer expectations. Recently the Nielson Norman group, a well-known usability think tank, reported that “20 percent of purchase failures in our studies could be attributed to incomplete or unclear product information.”

Many retailers assume that the most accurate and complete source of information about a product is the manufacturer or supplier. The unfortunate reality is that manufacturers are just beginning to improve their product information and data distribution process. Accepting information directly from these sources leads to factual errors, inadequate descriptions, poor images and incomplete attributes.

Retailers must realize that supplying the best, most accurate and complete information about a product not only increases conversion rate, but creates a key competitive advantage. In the same Salisfy study mentioned above, consumers cited not enough info on a product” as the number one reason they choose not to purchase a product online. This means that shoppers with an intention to buy online couldn’t find a source of good product information and ultimately turned to a brick and mortar store.

As the industry searches for a solution to this mounting problem, a growing number of retailers are taking a page from Amazon’s book, deploying an on-demand workforce of freelancers to help them address these issues at scale. In fact, 85 percent of companies in all industries say they intend to invest in growing their flexible workforce over the next three years.

While the demand for flexible talent is on the rise, so is the supply. In 2015, millennials became the largest segment of the American workforce. They desire much greater flexibility in their work than previous generations. More than 1 in 3 millennials report that they would take a pay cut if they could work in a location of their choice. More than 81 percent of millennials believe they should be able to set their own hours. This new outlook on work has radically changed expected tenure–with only 7 companies in the Fortune 500 maintaining employee tenure of 10+ years. In addition, 3.5 million American became eligible to retire in 2015. These Baby Boomer retirees are connected, digitally savvy and are joining the freelance workforce in record numbers.

As a result, the number of people trying freelance work is skyrocketing. More than 35 percent of Americans report doing some form of freelance work last year and more than 750,000 tried working independently for the first time in 2015.

While Amazon’s cloud infrastructure has changed the face of IT in the modern corporation, Amazon’s pioneering use of the Mechanical Turk workforce to improve their product data is just the beginning of the on-demand independent workforce revolution. In the future this always available resource will completely change how companies think about their resources.

Following Amazon’s lead, other companies have imagined a workforce capable of doing more than simple tasks and have built robust platforms that help companies build, manage and maximize their use of on-demand talent. Innovative retailers can solve problems that seemed insurmountable even a few years ago with unprecedented speed by leveraging this new workforce in the cloud.