The End of Omnichannel
This morning my friend Celine said in a meeting, “There will be no omnichannel, we’ll stop differentiating all the media we use and then it will just be retailing again.”
A smart observation and one that provides a good backdrop to discuss voice shopping.
A little bit of context
Personal interactions between shoppers and store associates defined retail for hundreds of years. In the 50’s self-service became the norm and from there ecommerce websites and apps have moved in to make the process of buying more efficient. But as these efficiencies have increased, personal interactions and the resulting conversations have disappeared, and as a result, a customer’s personal connection to a business has become less and less. This is important as business try to build relationships with their customers.
AI to the rescue
Thanks to new artificially intelligent, conversational interfaces, all that is about to change. It’s now possible for a customer to interact with their retailer, using their own voice and receive a fast, efficient, and best of all, personal experience.
We’re talking about voice shopping, conversational commerce, smart speakers, chatbots, and natural language processing – For retailers who want to build meaningful and long-term relationships with their customers, now’s the time to consider adding these technologies into their retail technology roadmap.
Chatbots first. Voice second
Voice still has a fair way to go before it’s going to overtake the chatbot “bubble” interface. Right now it’s great for one-way instructions like turning on lights, playing Jeopardy, or listening to a Rolling Stones playlist, but as requests get more complex, deep integration with the retailer’s systems, silos, product catalogs, MDM, titles, descriptions, items numbers, abbreviations and more becomes a pivotal part of creating a meaningful customer experience.
Search is a good way to demonstrate this: Let’s say we’re looking for an iPhone 8 on the Apple.com website. There are three variants of the product, the iPhone 8X, the iPhone 8 Plus and the iPhone 8. Assuming we know nothing about the products, visually the site presents us with a very effective side-by-side comparison and beautiful imagery.
Of course, if we asked a voice-based interface to do this simple search, it would become very complicated very quickly. Imagine being told the various details of each phone as you try to compare them.
Now imagine something more complex like searching Amazon for a pair of hiking boots. This produces hundreds of results – easily scrollable on a screen, but next to impossible for a normal human to even remember past the first five items as the voice agent describes each product.
A text-based chatbot has similar issues, however, it can show images in a carousel. For voice, this problem can be solved using search refinement techniques, where the bot can ask some qualifying questions before it starts to talk about a product. Still, even though today’s product catalogs may be attribute rich, the average voice agent may not know which are the most important attributes to use in those qualifying questions.
Building product search for both text and voice interfaces requires strong A.I. to disambiguate the request, then requires well marked-up product data. A different breed of A.I. is required for this, one that imbues semantic meaning product data based on what people search for and what they finally select. Ecommerce labs are steadily improving the accuracy of these technologies.
Implementing chatbots today will make voice easier tomorrow
Because the frameworks of the conversational interface for a chatbot and a voice device are effectively the same, retailers, who offer chatbots to their customers now put themselves in a good position to make a successful leap to voice shopping once customers adopt this new way to shop.
Conversational commerce is a reality now, and voice-shopping is not far behind. Interested in how you can add chatbots and voice to your omnichannel roadmap?
Contact us today; we’d love to hear what you’re planning.