Op-Ed | Chatbots Won't Solve Everything

NEW YORK, United States — "Conversational commerce" is the idea that mobile chat interfaces, such as Facebook Messenger, WhatsApp and WeChat, could unlock a new frontier in shopping. Chat bots — text-based automated software applications — are at the centre of the discussion, promising to help companies efficiently interact with the millions of active users on messaging platforms, without requiring vast human resources. Chat bots are all the rage, especially after Facebook's recent announcement that it would open up Messenger to businesses and the release of Kik's bot platform. But to make "conversational commerce" a reality, it's critical to be clear about where chat bots can — and can't — help.

So far, most bot-powered chat interfaces fail to answer a fundamental question: how do they provide actual value for the customer? For example, a chat bot that takes minutes to deliver the weather forecast, which can be scanned in seconds by looking at a screen, isn’t helpful. So where can chatbots add real value — specifically in the fashion industry?

As with any new product, businesses should consider chat bots an addition to their toolkit, not a substitute. Chat bots should add value for customers, not move existing experiences to a new interface just for the sake of it. More specifically, adopting the lens of a conversion funnel — where consumers move from awareness to consideration to conversion — helps narrow down the use cases for chat bots. At a high level, I’m betting that chat bots will be most effective for awareness and consideration — not conversion — and that the best businesses will take a hybrid approach, blending bots with human-to-human interactions.

As businesses experiment with bots and interfaces, the best way to stay on course is to constantly ask yourself one question: How does this help the customer experience?

Bot-Powered Product Discovery and Awareness

Endless sales and markdowns are endemic to fashion, caused in part by the failure to accurately match supply and demand. This isn't surprising considering most fashion businesses design and buy with more gut than data. Additionally, the standardised format of stores means brands struggle to get the right products in front of the right people. Even online, where it’s possible to provide a personalised shopping experience, there are four problems 1) generally speaking, everyone sees the same products; 2) items are sorted categorically; 3) the product selection is seemingly endless; and 4) the consumer has to actively visit the website to browse and purchase. Even for the most loyal customer, repeat visits to e-commerce websites still feel like visiting a store for the first time. An alarming number of e-commerce websites (and their newsletters) still aren’t tailored to a customer's gender, price point or favourite styles.

Can Chatbots Help?

Before diving into chat bots, it’s important to step back and consider how people find and receive data. People generally get information in two ways: either someone seeks it out, known as “pulling” information, or the information comes to the person, known as “pushing” information. Either you seek out a product (pull) or a brand emails you about it (push). Early chat bots have employed a pull model, meaning the customer has to actively work to get results. As a consequence, many users have reported sub-par experiences. A push model, where a company pushes information to the customer, is much more compelling. Supreme, for example, sends a push notification through their app every Thursday when new products drop, enabling the brand to quickly turn a passive customer into an active shopper.

Combining a personalisation engine with push notifications might unlock the potential of chat bots. Each customer has a unique set of interests, price points and styles. Brands should use a consumer’s purchase history and other preferences to deliver a personaliaed shopping experience. With messaging apps, this becomes more feasible, as brands have a direct line to the device where consumers spend the most time — their phones — and the open rates on texts and push notifications are significantly higher than mass emails and retargeted ads.

Personalised push notifications will logically become bot-powered. Brands with excellent data operations will be able to build a new layer of notifications, using chat bots to send personalised recommendations that can help build awareness at the top of the conversion funnel. Customers don’t like getting irrelevant messages, but getting the right message at the right time is valuable.

A Hybrid Approach to Consideration and Conversion

When moving down the funnel from awareness to consideration and conversion, a hybrid approach that combines chat bots with humans makes the most sense. Below is a theoretical example of an interaction between a user, a chat bot and a sales associate:

A user gets a bot-powered message from a brand, say Acne, with three products (awareness via bot). The user asks to learn more about one product and whether the SoHo, New York store has her size (consideration via bot). The bot then checks the stock of the SoHo store, confirming the item is in stock. The user replies, asking if she should size up or down. The message is passed to a sales associate, who answers her query and also recommends two other products in stock in her size (consideration via sales associate). The user decides to either reserve the two products to pick up in store or have them shipped to her home (conversion via bot or sales associate).

Early experiments that take a full-bot approach to conversion have been rigid and unsatisfying for three reasons: talking to a robot is 1) far from perfect; 2) slow; and 3) buying is a process, not something you force — the funnel exists for a reason. In the future, artificial intelligence will surely help to tackle the first two problems. But today, introducing chat bots and interfaces in the right scenarios, where they stand to add real value, is key to making them impactful.

Supporting Styling and Customer Service

A hybrid bot-plus-human approach also applies to styling and customer service, where it would certainly improve the status quo. The relationship between shoppers and luxury brands has always been personal, across many touch points, with sales associates commonly giving clients their own phone numbers. Chat interfaces will allow companies to scale and democratise these kinds of relationships, allowing a wider number of businesses to offer a higher level of service.

Interestingly, this might call for many more buyers and stylists, since they will have infinitely more personas to shop for. The days of a single store shopping for “our girl” are over. It’s “our girls” and there will be hundreds and thousands of them. Now, aided by deep reams of data, buyers can make informed decisions for each person, while also delivering better, more specific results. When paired with a conversational interface and possibly a bot, the associate-client relationship can be both intimate and scaleable.

The hybrid approach also offers a direct line to a brand’s customer service team, which is a simple yet powerful feature. No more lengthy calls to phone banks and tracking down receipts — just send a message to the customer service team and someone will be in touch shortly. The customer service rep will know your previous orders and product browsing history. And bot messages, such as order and shipping confirmation notices, will exist in the same channel as the human conversation, creating a complete, searchable record of all interactions between a customer and a brand.

If used incorrectly, chat bots and interfaces run the risk of making the consumer experience worse. They simply will not solve every problem. As businesses experiment with bots and interfaces, the best way to stay on course is to constantly ask yourself one question: How does this help the customer experience? If the answer is hard to come by, maybe there's more work to do.

Richie Siegel is the head of NewStore Labs.

The views expressed in Op-Ed pieces are those of the author and do not necessarily reflect the views of The Business of Fashion.