Select Page
LiveArea-conversational-chatbox

A Conversation About Conversational Commerce and Chatbots.

BLOG

Authors: Soumya Kundu and Sumegha Mantri

August 17, 2021

On a quiet Thursday afternoon in a COVID lockdown city, Sumegha was busy replying to emails at her temporary home office setup, as her speaker played her favorite Spotify playlist. Her mobile notified her that a parcel was being dropped off soon at her doorstep. 

The doorbell rings, Sumegha walks across the living room to open the door. A quick swipe of her hand, a form was exchanged with the box of flowers that she had ordered via WhatsApp. As she closed the door and returned to her seat, a notification from her colleague buzzed on her screen.

 

Shomu: Hey Su, we are doing this piece on the state of conversational commerce, want to join in on a brainstorming session?

 

Su: The timing could not have been more perfect, just last week, I ordered a cake and started a flower subscription through chatting. One was on WhatsApp and the other on Instagram – one was a real person and the other was a chatbot. Got me thinking about conversational commerce experiences…

 

Shomu: Cool! That sounds interesting. How did the conversation go? Did the conversation with the bot sound mechanical (beep bop noises)?

 

Su: You know, the one with the chatbot was rather odd – it kept repeating the same loop of the menu-selection and got stuck at customization. It felt like I was going round and round in a maze with no end in sight!

But the one on Instagram with the real person on the other end was smooth. She asked me how I wanted the cake customized, promptly checked with the chef, and then got back to me. That was pretty smooth, and great customer service I must say.

 

Shomu: I’m curious how you got to the chatbot in the first place?

 

Su: I was on their website and I wanted to know if their subscription model was flexible. I saw a WhatsApp icon and thought it wouldn’t hurt to ask… that’s how it started. Also, I must mention that it was quite late in the evening, so I couldn’t really call and speak directly to the customer service team.

You know what, chatbots are convenient. Imagine having a person available 24/7 – that would mean a substantial cost for the business. At least I could sign up to the subscription at an unearthly hour of 1 am without bothering anyone.

 

Shomu: True! But while the Conversational AI/ Chatbot brings back the ease of conversation and reactivity to the ecommerce experience, it still has a long way to go in understanding the nuances of human conversation. At the moment they are well suited for repetitive pattern-oriented task like support services, reminders, etc.

I think this also works well for people who are not comfortable around sales associates hounding them, or introverts. It’s like the chatbot becomes a companion who is on standby, and if needed it would get help. For sensitive issues it could also then connect us to a real person.

 

Su: Indeed, it’s a calmer commerce if you like it. It’s there when you need it and vanishes in the background when you don’t. Quite contrary to the present nature of in-your-face, always asking for your attention commerce.

 

Shomu: Yes, it’s very off-putting, my email and SMS messenger apps are filled with these attention grabbing, jarring messages from all the different brands’ bots. What I also find extremely annoying is having content thrown at you without context.

 

Su: I feel you, but at least this one talks back, is contextual, and actually useful! But yeah, the technology will evolve and get better you know…

 

Shomu: Agreed. From the merchant perspective, having efficient chatbots would mean saving a substantial amount of money on the customer service team. This could help them reduce human error and enable efficient customer service with minimal resource costs.

 

Su: Hey! Do you know more about how chatbots are made? I mean, what goes into making one?

 

Shomu: It all starts with mapping the proper intent, so there are rule-based chatbots and AI-based chatbots. Rule-based chatbots follow a very strict if-then-else format. Whereas the AI-based chatbots try to map the proper intent of the customer input and assign proper responses accordingly.

The rule based chatbots are very straight forward, like ordering a pizza. When you order a pizza from your favorite pizzeria, an operator asks for your order on the phone. The chatbot can do the same, asking you about the size of the pizza, the type of crust, topping and amount of cheese. It will then request your address and a payment method. These are logical, pre-determined steps, that only require you to click through them.

 

Su: And I guess some amount of the magic is not only in the speech, but also in the “umm, hmm” and the pauses. Those are the real nuances of conversation that make it humane.

 

Shomu: That’s where the AI enabled smart chatbots come in. They are designed to simulate near human experience. They can have free flowing conversations and understand the language, intent, and sentiments.

 

Su: Ummm, like in the movie, Her? Wouldn’t it be fantastic if we could reach that level of natural conversation with machines! But honestly, my present experience with chatbots was far from it, it was pretty confusing.

 

Shomu: Yup, that’s because they are hard to build and require a lot of data sets to learn and train from – like Siri or Alexa.

You know what? This reminds me of another recent observation, which might be a good use case for conversational interfaces. In the train station, they had installed kiosks with touchscreen interfaces to help book tickets faster and ditch the queue. Although it was good, one thing I noticed was the audience for whom it was meant, the daily travelers, mostly minimum wage workers who travel to other cities, were not able to use the facility. I noticed that when they saw someone interacting with the kiosk, they would immediately approach them for help getting the ticket. That made me think that in future these kiosk experiences could come with an assistant feature. That would help gain trust of any new users who may not be digitally literate.

 

Su: Oh! You mean like avatars on a screen, trained to help? That’s a great idea. You know I had a similar experience at an ATM – a person who couldn’t read or write was waiting for a long time for someone to come by who he could ask for help. I ended up translating for him and helping him make the transaction. A voice interface in the local language would have been very helpful. Unlike visual digital interfaces (which still need people to be literate and come with a slight learning curve), most people learn languages and local nuances in early childhood. This makes conversations and particularly voice-forward interfaces more intuitive and inclusive in nature.

 

Shomu: Oh yes, and just think about the advancement of the haptic technology – this could mean we can create better inclusive experiences for, say, the visually impaired. The application could be in airports, malls, parking lots, elevators, cinemas, hospital, supermarkets…

 

Su: Yeah! I guess moving forward, the future lies in holistic experiences that feel natural. Reminds me of what the computer scientist Mark Weiser said, “The experience of using technology should be as refreshing as taking a walk in the woods!”

There is also another advantage of conversation and voice interfaces – they are there when you need and when you want to engage.

 

Shomu: Yes, they give us, the users, a choice of control. A very likeable feature in the age where our attention is always being pulled in so many directions.

All in all, the conversational aspects of chatbots have come a long way, but it still has a lot of avenues for further exploration with the integration of newer technologies. For now, I am happy if the chatbot is intelligent enough to direct me to a fellow human when it cannot provide me with a solution instead of sending me in an endless loop

 

Su: True. So, what about the brainstorm session on conversational commerce?

 

Shomu: Oh I think we discussed a lot of good points. You know what, I will transcribe our discussion as is and share it with the team and see what they have to say. Thanks, Su!

 

Authors:

Soumya Kundu, User Experience Designer and Sumegha Mantri, Associate User Experience Lead

RELATED ARTICLES

RELATED ARTICLES

Understanding Intent

Understanding Intent

As shoppers become more digitally savvy, and businesses lean more into technology to differentiate themselves, eCommerce teams are often tasked with both creating fresh brand experiences and refining the ones already in place to keep up with high demands of omnichannel shoppers.

CONTACT US FOR MORE INFORMATION

gradient bar
Dallas    |    Los Angeles     |    New York    |    Raleigh    |    Seattle    |    Bangalore    |    Liège    |    London    |    Sofia

NORTH AMERICA    1.800.920.4959    |    EMEA    +44 (0) 20 3475 4000

NORTH AMERICA
1.800.920.4959

EMEA
+44 (0) 20 3475 4000

CAREERS     |     CONTACT US     |     SUBSCRIBE