From books to groceries, Amazon.com has disrupted the way consumers shop. But while much of the technology used to facilitate online shopping is novel, many of its ideas came from observing what customers want.
For instance, the US tech behemoth launched
Amazon Go, an experimental cashier-less convenience store in Seattle late last year. The idea was born out of the observation that queues in grocery stores can be frustrating, says Paul Misener, vice-president of global innovation policy and communications at Amazon.com INC.
“Think about it: If you go to a grocery store, you pick out the things you want to buy, you are ready to go and then you have to wait and wait and wait to give them money. You’re surrendering your money to them, but you have to wait to do it! That’s a terrible customer experience, so we figured that by installing automation in stores that will automatically detect who the person is, what they buy, what they examine, put back and walk out with, it will improve the customer experience,” Misener says.
The “Just Walk Out Technology” used to sense customers is similar to what is used in self-driving cars, such as computer vision, sensor fusion and deep learning. Customers will be charged on their Amazon account automatically after leaving the store.
The principle is called customer obsession, which means that the company focuses more on observing what their customers want rather than what their competitors are doing. From there, Amazon turns the observations and ideas into products by working backwards.
Misener explains what this approach entails and what it means: “So, today we will write the press release for the product or service, together with an FAQ that will be released sometime in the future. That does three things for us: it moralises the decision to work on this new product or service, so we know what we’ve committed to do and we can’t deviate from it if things get hard. The second is that it’s in customer terms, because it’s a press release, it doesn’t describe how we’re going to do it. Lastly, it’s an acknowledgement that it’s impossible to do what we want today. We’re saying we understand that we’re going to have to invest and innovate all the way to the point of delivery.”
Take Amazon Prime Air, for example. Misener say the press release the team wrote for it only specified that the goal was to make the delivery within 30 minutes of an order.
“It wasn’t like you needed to design an autonomous drone that has quiet propellers and a computer vision systems. It was just using unmanned aerial vehicles to make deliveries within 30 minutes,” he says. Prime Air delivery went through a private trial in the UK last year and Amazon is still testing the service in many countries.
Attempting to observe what customers want and ambitiously setting goals can be a risky proposition with a potential for costly failures. While much attention is paid to Amazon’s successes, the online retail giant also has had its share of failed products. One example is Amazon Fire, the smartphone released in 2014 that received an underwhelming response.
“We’re willing to experiment and fail, and that willingness to fail is part of being innovative. There is a reason I carry around an iPhone - we have a mobile phone product that didn’t work out so well and customers didn’t flock to it, so we learnt something. We have incorporated some of those technologies into other Internet of Things (IoT) devices,” Misener says.
The Amazon Marketplace was also the result of another failed product. Early on, Amazon decided to allow third parties to sell their products on its website like an auction website.
“But the investment community went crazy! They were like ‘Why are you doing this? You’re a retailer, you’re buying stuff wholesale and selling it off at retail; why would you let somebody come into your website and compete with you?’ It turns out the auction business failed terribly and nobody used it. Seriously, it was bad,” Misener admits.
The company then decided to introduce zShops, which is a separate corner of the Amazon website for third parties to sell their goods. Again, the experiment failed because no one clicked on the website.
“Then we finally got it right by introducing what we call our Marketplace, a single detail page and that is what you see today on Amazon’s website. You see a product, but you also see a variety of sellers, and the one who gets the sale is going to be the one who offers the best value, and oftentimes, that’s not our retail business,” he says.
Sometimes, innovation also means that old products may be cannibalised to make way for new products. Amazon’s Dash button, which allows customers to order refills for goods such as detergents with the click of a physical button, was originally ridiculed by many as an absurd idea. Some brands that customers can use with the Dash button are Colgate, detergent Tide and Clorox disinfecting wipes. Now, the function can be replaced by Amazon’s cloud-based voice service, Alexa.
“I think it was an innovative idea that we deployed and is enjoying some success. Is it the case that we have to stop with that and not move on to Alexa-enabled devices? Of course not. They can operate in tandem, there can be places in the consumer’s home or office where both things can coexist, so it’s all about giving customers a choice,” he says.
Misener says despite the popularity of e-commerce, there is still a lot of room for Amazon to grow. The company is building more warehouses and investing in technology, such as its 2012 acquisition of Kiva Systems, which makes IoT enabled robots, to speed up the logistics of delivery.
“It’s especially important to recognise that here in Malaysia, e-commerce accounts for about 5% of sales while in the US, it’s 9.5% of retail. That’s not even double digits. So, in both countries and in the world, there is a tremendous amount of headroom for growth and innovation,” he says.
Amazon launched Prime Now - a two-hour delivery service - in Singapore last July, and released the full Prime membership service in December. Singapore is the company’s first market in Southeast Asia, but its expansion plans may face fierce competition from Asia-based e-commerce players such as Lazada Group, which is owned by China’s Alibaba Group Holding Ltd.
Misener says in this region, its cloud computing service, Amazon Web Services (AWS), is more popular than its online retail service because of a stronger demand for it.
“The biggest frustration is that we can’t possibly do everything at once everywhere. It’s frustrating to me that everything is not yet here in Southeast Asia, but cloud computing is, fortunately. We’ve deployed different businesses first in different countries over the years. Traditionally, it has always been retail first, but in Brazil, for example, it was digital first, so Kindle was there before retail. In this region, it was AWS, and I think it’s because there are entrepreneurs and small businesses here who could benefit from it,” he says.
The biggest technological innovations going forward will be in automation and voice-powered technologies such as Alexa, Misener says. The system is programmed to rely on machine learning and artificial intelligence (AI) to pick up on the owner’s preferences and language. Amazon has a development centre in Cambridge, the UK, that is working on creating natural language recognition in Alexa. The centre was launched last November for the research of products such as Alexa and Prime Air.
“The concept is so that Alexa doesn’t just take words and stick them into a word search, but it actually tries to figure out what you mean, not just what you say. [For example,] if I wanted this phone to be removed from the table and taken outside, there are 50 different ways I can say that. In every instance, it is very clear to a human that I want the phone outside, but to a machine, it can have very different answers, depending on how I say it,” he says.
Going forward, AI and machine learning will disrupt traditional industries such as agriculture in the same way, Misener adds.
Amazon has another lab in Europe that is working on computer vision for the delivery drones. The technology will help the autonomous drones detect where to land, for instance, to ensure the safety of people and objects it flies over. Misener says Amazon is working with regulatory authorities to demonstrate the safety of its drones.
“It is fair to say that by innovating, you do create regulatory cases of first impression. Regulators haven’t had to think about the operation of autonomous, low-altitude drones until enterprises like ours start to develop them. We’re in a funny spot because we’re just about the only enterprise I can think of that is working both as the developer of the technology and the user of it,” he says.
Misener’s advice to entrepreneurs is to focus on creating something that differentiates their business and use existing technology such as AWS where it is available so that they do not need to buy legacy equipment.
“Don’t try to reinvent stuff that already exists. Focus on what differentiates your business as an entrepreneur and rely on other service providers to do the stuff that have already been done. For example, if you’re a seller, why would you go and invest in your own website? Just sell on Amazon. Just use the services and focus on making your product a better idea,” he says.