Below, is the fourth product in my 100 Products, 100 User Journey’s project. Meet Amazon Personal Shopper, a personal shopping service provided to Amazon customers.
While the concept of a personal shopper is not a new one, Amazon Personal Shopper (APS) solves a problem for the customer in the short term and provides a key transition for Amazon’s long-term artificial intelligence (AI) strategy.
Why a Personal Shopper?
Despite the ability to sort by pricing, refine search results, and get a product in two days with Amazon, there still exist major customer pain points in the purchase funnel
Let’s take purchasing a mattress for example.
The Pain Points in Shopping for a Mattress Online
- Choosing is hard: You search “mattress” on Amazon and there are 50 options, all of which have high ratings and a lot of reviews.
- Topic specific knowledge: What’s the difference between a gel foam, memory foam, and latex mattress?
- Product Quality (General): You want to know more about a product’s general quality (i.e. longevity) and have +100 comments to sort through.
- Product Quality (Specific to Me): You want to know about whether the product meets your specific needs (i.e. firmness) and have +100 comments to sort through. You don’t know if another person’s definition of “firmness” is the same as yours.
- Picture Discovery & Availability: You can’t see or touch the product in person so you need lots of photos to give you more views of the product. Photos are not always available or easy to find.
In an ideal world, Amazon Personal Shopper would be solved with a well-developed AI algorithm that could precisely pinpoint the best mattress for you based on data collected from millions of similar transactions with other customers. You would tell Amazon Personal Shopper what you were looking for and Amazon would automatically know what product to provide you.
However, this isn’t possible with current AI technology. A good intermediary strategy would be to hire independent contractors who could help customers select the best products for themselves. Amazon Personal Shopper would help Amazon position itself at the top of the purchase funnel where shopper ideation occurs, rather than in the middle of the funnel where a shopper already knows what they want. This enables Amazon to increase the volumes of customers coming in at the top of the funnel and grow their basket size as they make their way down the funnel.
A personal shopper solves the customer pain points by providing accurate, specific, and directional advice on what product to buy.
Amazon customers can review independent contractors based on how well the personal shopper selected products that fit their needs.
Implications for Amazon’s AI Strategy:
Perhaps the most important part of Amazon Personal Shopper (APS) is the data it would provide to Amazon to inform natural language artificial intelligence. Microsoft has access to copious amounts of natural language text through channels like search and emails to influence Cortana, their AI bot. Google Assistant has access to similar information to inform its interactions with customers. Amazon lacks this data and Amazon Personal Shopper would provide much needed natural language text by tracking conversations with actual customers. Data could include:
- Vocabulary used during various points in the purchase funnel
- Speed to final purchase decision (length of conversation/ conversion)
- Review scores
- Product return rates in APS transactions vs. regular transactions
Amazon could use this data to eventually train Alexa or some other AI bot to accurately predict what shoppers are going to want and buy.