Today, you walk into a brick-and-mortar retail store. No, this is not a joke, and that’s not the punchline. E-commerce continues to take bigger bites out of the retail pie, but there are still unrealized milestones for the innovations and convenience of brick-and-mortar. One of the last frontiers is the presence of the friendly, helpful shop assistant.
How can I help you? is something we’re used to hearing when we walk into a store, and we’ve taken it for granted. For every intuitively laid-out eCommerce website that includes UI-friendly features such as AI auto-complete or breadcrumb search results, there are many more that, to be blunt, completely suck. It would be a godsend to have a friendly shop assistant pop up and ask a few simple questions about what I’m looking for. Can it be done online? This article will look at available possibilities and share some tools, tips, and advice.
How to piece together your own personal assistant
While virtual shopping assistants are in development, a program that will feel human to your customers is not quite within reach — or within budget. However, it’s not too hard to combine several different applications to give your visitors a taste of the best features of a shopping assistant without too much of a splurge.
Chatbots are nothing new, but their UX has improved and their applications have diversified. These days, it’s easy to get creative with integrating chatbots into your operations.
Facebook Messages: You know your customers are scrolling through their Facebook feed half the day; why make them leave the application when they want something from you? Having an easily accessible ordering system is like having an on-call personal assistant — and instead of navigating to your website, messaging you on Facebook makes it feel more like they’re talking to a human. Sephora has really been leading the charge to the future in the beauty world with two different chatbot features within Facebook Messenger using Assi.st: Customers can message them to set up an appointment with a beauty consultant, or they can get advice on purchasing decisions.
Ordering food for pickup or delivery has also taken off in the Facebook Messenger world. Starbucks is just a few messages away from being available to pick up at your local shop, Dominos can tell you the daily pizza deal, and Pizza Hut lets you complete the entire ordering experience without even leaving Facebook. These are all done using various chatbots with the same experience as when you chat with a friend.
Customer Service Chatbots
Using chatbots to help your customers with customer service questions is like having a virtual personal assistant who doesn’t sleep. They won’t be able to handle the big stuff, but automating the little stuff can take a weight off your bottom line’s shoulders. Aptly named, a service like ChatBot can be used to easily build out your scenarios, questions, and actions — not quite Bandersnatch levels of complexity, but it gets the job done. It has a high rate of return, too: In a test, a chatbot was able to resolve 82% of interactions without the need for a human agent.
MongoDB has a customer service chatbot like this that can ascertain if a visitor is a qualified lead by asking a few questions and if they are, direct them to the correct sales rep. Sephora makes another appearance in this arena — are you surprised they’re in the chatbot customer service game, too? On their website, not only can you ask basic questions, but you can even get makeup recommendations from their AI. Customers can scan a photo of a makeup look they like from anywhere and get advice on what to get to cop the look.
Convincing your visitors to get emails from you isn’t an easy task — what if a chatbot could convince them for you, and only send them exactly what they want to see? That’s what TechCrunch’s bot claims to do, without any extra effort on the part of the subscriber at all. When the reader signs up for personalized news using the chatbot service, its AI software then keeps track of the type of news that they read and sends them only articles that it thinks they’d be interested in.
Let StitchFix try to know you better than you know yourself
Building it into your business model
Wouldn’t it be great if your customers always felt like they were receiving personalized assistance from you? There are a few companies and industries that have managed to build the feel of a personalized assistant into their business model.
Part of the equation of a successful subscription box is finding out what your customers like to send them the right thing. Stitchfix’s model centers entirely on getting customers to tell Stitchfix what they like so Stitchfix can send them things they might like. This personalization feels extremely unique, as each person is paired with a personal stylist after filling out a hefty detailed quiz. The customers pay a subscription fee, which is deducted if they keep at least one of the items sent to them.
However, no business could profit from personal stylists looking over each individual profile and sorting through a massive catalog of items. Humans are terrible at quickly and efficiently processing large amounts of data and making decisions — that’s a job for artificial intelligence. AI is how Stitchfix efficiently scales up, with its algorithm looking at trends, measurements, feedback, and preferences to narrow down a list of suggestions for the stylist. The AI assists the stylist, who assists the customer in true tech-human harmony.
If You Liked That, You Might Like…
A true personal stylist knows what you like and what you’ve bought and uses that information to suggest other things you might like. It’s not hard for artificial intelligence to mimic the “if you liked that, you might like this” personalized suggestions. Half the battle is getting customers to sign up so you can collect their data, and the other half is effectively using that data. Who does a great job of this? You guessed it. Amazon.
Amazon knows that 60% of the time, someone looking at a Keurig coffee maker has also looked at disposable K-Cups and probably actual cups to drink the coffee. What does the AI do? Suggests those products to everyone looking at a Keurig. It’s kind of like having a virtual assistant who’s constantly trying to guess what you want based on what you’ve searched for, what you’re clicking on, and what millions and millions of other people have done in your situation.
Can AI help you find your perfect product?
Looking to the future
Researchers and developers are always trying to answer the question: Can we make a truly personal virtual shopping assistant? For now, there are two interesting applications that get pretty close.
One is Macy’s On-Call, which was surprisingly ahead of its time, and also uniquely combines AI and virtual shopping assistant features with visiting a brick-and-mortar store. When customers visit a Macy’s store, they can hop on their phone and access the On Call function to ask questions about inventory, an order they’ve placed, or even get directions to the location of another department. All they have to do is type in questions and they get responses instantaneously.
Macy’s On-Call was tested out in 10 stores, but hasn’t progressed much beyond there. However, it did seem promising, and they partnered with IBM Watson. Due to the rising popularity of using chatbots, it’s an investment that may pay off for them in the future, and is worth attempting to emulate for a virtual eCommerce store.
However, the latest and greatest development is an app called Elly. Elly is the closest existing thing to a truly smart virtual shopping assistant — however, she’s still in the developmental stages. She’s an AI that helps customers find their perfect product through asking a series of questions, balancing features, price, and anything else the customer says they care about. She’s in the testing stages at the moment, but you can currently enlist her help with finding your perfect smartphone if you want a taste of the future.
How can I help you?
A personal assistant knows their business inside and out. They also aim to know as much relevant information about their customer as possible, to help them make smart purchasing decisions and leave satisfied (and, of course, return for more). Finally, they want this to happen in a natural and efficient way.
The problem with using human personal assistants is that they can’t scale up efficiently and use large amounts of data in a meaningful way. The future of virtual shopping assistants is to combine the helpfulness and personalization of a human assistant with the data-crunching power and speed of artificial intelligence. A single application can’t do it all (yet), but combining a few tools that are available now potentially unlock new levels of efficiency for eCommerce businesses.