Will chatbots be successful and part of our future?

Recently, R/GA San Francisco kicked off an Accelerator program focused on defining the future of commerce — from messaging chatbots, to visual recognition and connected experience technologies, to innovations in fulfillment, returns and staff training. As we work with the startups to help define their business and brand, the conversation always seems to gravitate towards chatbots, artificial intelligence, natural language processing and machine learning.

But will these smart chatbots live up to the hype? Do they have the staying power to become part of our everyday lives?

New technologies often fail the first time around, either because expected basic features are not ready, or because they don’t clearly identify how you can use them on an ongoing basis.

Take Google Glass, for example. People were excited about its release. You had easy access to apps via voice commands and were able to record videos of the world around you. The main problems with Glass were that it was too expensive, and people didn’t really know why they needed it. In the end, people wearing Glass became known as “Glassholes”.

Apple Watch had a similar outcome. While many Apple fans were excited about wearing a phone-like device on their wrist, many others soon realized that the Watch lacked any meaningful apps, and was therefore quite useless.

And then there was Microsoft Tay, who was pretty impressive. Right up until he started making racist comments.

While I think many chatbots will fail, I believe that many have the chance of succeeding — IF we can make them useful and relevant. So how do we do that? The answer is simple.

I suggest we build bots that do the things that robots do best, and not build them to do things that humans can do better. Things like quick calculations and looking up data and finding the best answer from millions in a matter of seconds. This is what bots are made for.

Like in customer service. There are many things that chatbots can do better than humans can in this area. Bots can answer faster than humans. They can memorize every tech spec of millions of products, or recommend recipes based on only a few inputs. Why would a company want to spend time and money on training a customer service agent to respond to simple FAQs, when these questions could be answered by a bot using NLP (Natural Processing Language)?

Great customer service is one of the reasons companies sell more in retail stores than they do online. According to a study, 90% of customers are more likely to make a purchase when they receive assistance from a knowledgeable associate. Problem is that companies cannot scale agents online the way they do in retail stores.

That’s where companies like Agent Q come in. By combining AI with product information and user preferences, Agent Q bots can answer more than 25% of common customer questions. This, combined with their plug-and-play dashboard that helps real customer service agents answer questions quicker, puts Agent Q in a unique position to be able to help businesses provide better, faster customer support and save them a ton of money in the process.

Customer service is just one fertile territory for chatbots, one we should be exploring as we work with our commerce start-ups and our clients across the network.

But what other skills do you think chatbots would be better at than humans? Healthcare? Policing? Disaster relief? Politics? I’d love to hear your suggestions in the comments section below.