Help Wanted: 1 Million Personal Shoppers

Digital commerce is at a challenging point. Over the last few years, actual year over year percentage growth has decreased, according to the US Census Bureau*. We are now reaching a point where what was once seen as a benefit – endless aisles and order from your desk – is now seen as a burden of too much choice and a need for a simpler and easier experience. While the eCommerce world wrestles with the migration from a desktop environment to a smaller and more distracted smartphone user, (almost 50% of traffic in the 2015 holiday season came from mobile devices) the user is stuck trying to understand why shopping online seems almost antiquated and challenging compared to so many apps and intelligent digital experiences that respond to them as an individual. eCommerce needs some kind of digital personal shopper.

While there are many things to dislike about the digital shopping experience, there are several benefits. You can visually curate what you like and don’t like, products are merchandised by the season and usually by your geography and when push comes to shove, there is usually a sales associate that can help locate what you are looking for be it size, style or color. Conversely, online, most of the products are hidden in digital aisles (pages), most sites treat their visitors somewhat generically and promote popular items regardless of where you live, and finding something similar or that has elements of what you like, can be incredibly frustrating and challenging. Essentially, eCommerce sites are little more than digital catalogs that ignore the individual intent of the shopper.

When I land on an eCommerce site and start clicking on things, I want the site to respond to my expressed interests. It should guide me to merchandise that’s similar to the preferences I’ve been indicating. When I use site search, I want the results to be presented so that the items that make the most sense for what I have been looking for are stacked at the top (a common color, style or even gender for goodness sake). When I’m on a smartphone I want the site to have an ounce of respect that my 5 inch diagonal screen is not the best environment to show me a mile long page of random 100 results. I also don’t want to be put into a bucket that shows me what other people like me looked at. I am an individual and I want online stores to respond to what I have shared through my shopping behavior.

Force feeding me a generic shopping experience that’s built for a segment of 5,000 people isn’t going to work. I don’t want a “sort of me” experience.

Basically, I want an intelligent personal shopper who can respond to me quickly, help me find what I am looking for and show me items I may not know about.

But the question is how can you have 1 million plus digital personal shoppers waiting on the line as each shopper hits the eCommerce site? How can those digital personal shoppers respond to each shopper’s individual intent in real time with super relevant merchandise when they have to sort through a huge catalog of products? How can those same digital personal shoppers make sure they are retaining the shopper’s past experience and preferences so they can be leveraged for the current visit? In short they can’t. Instead, eCommerce needs a way to programmatically address that volume, so it meets the real time challenge.

Other sectors of the digital space are already doing this. Most notably, the digital advertising space responds to clicks in seconds and optimizes the revenue of each click by presenting the most profitable (read relevant and most likely to be clicked) ads to each individual user. As a matter of fact, one of Google’s biggest inflection profit points was when they moved from using collaborative filtering to optimize digital ad response (think segmentation) to individually relevant digital as response.

For digital commerce, the definition of success is slightly different but the application is very similar. How do you put the right product in front of the right person to create a click or even a successful event? By applying intelligent machine learning algorithms to this challenge, companies are essentially creating digital personal shoppers. For eCommerce you need to make sure to put in merchandising rules that are typically unique for each retailer. But within that rule set, eCommerce sites can now present the most relevant products to each individual user by understanding the attributes of the products shoppers are looking at (material, brand, color, gender, category, product, length, etc.) combined with the attributes of the individual shopper (location, past products viewed and their attributes, device, time of day, etc.). The secret here is to look at all of the visitors shopping behavior, each click, site search, product viewed, not just items that are purchased. Not only does that make the digital personal shopper concept more responsive to an individual as they shop, it also makes sure that the system works for anonymous visitors who have never purchased anything from the retailer.

A handful of eCommerce sites are already doing this incredibly well. Part of the secret is in the fact that they are doing it implicitly, not explicitly. This makes the digital store intelligently responsive to things the shopper might not even consciously know they prefer. A standard color, material or brand are suddenly a correlated preference that help the visitor get to the products they like faster, essentially clarifying the signal through the noise. Digital stores do this by prioritizing search results to the individual’s preferences, allowing category pages to present the most relevant products first, inserting predictive merchandising widgets on home pages and product pages and inserting the items most predicted to be purchased by each individual into email content. These aren’t things that the user looked at before, these are items that are brought forward without being viewed previously. This is a real digital personal shopper that brings the customer things that meet and exceed their expectations without being asked. And when this happens effectively, success for the shopper comes in the form of purchasing things more often, and for the retailer their conversion rates increase dramatically. It turns eCommerce into a shopping experience that involves discovery and surprise.

I guess we can remove that help wanted sign. The positions have been filled.

This article was also published in Innovative Retail Technologies

Photo: Flickr / Polycart

Kurt Heinemann
Kurt Heinemann

Kurt is Reflektion’s CMO. His resume includes CMO positions at Marketwired and Monetate. He’s also held senior executive positions at Priceline, Time Warner and Walker Digital Companies. Despite living in Yankees territory, Kurt is a die hard Tigers fan.