Organizing Conversational Services
It seems like every day there's a new Assistant-for-X service, messaging-based business, or AI-powered chatbot. It can be difficult to keep up, so we wanted to put together a clean overview and maintainable list. You can access the raw data and make suggestions at GitHub.com/ArgoHQ/List.
In defining our criteria for inclusion, we began by looking at Chris Messina's foundational article on Conversational Commerce, where he offers the following definition: [it] largely pertains to utilizing chat, messaging, or other natural language interfaces (i.e. voice) to interact with people, brands, or services and bots that heretofore have had no real place in the bidirectional, asynchronous messaging context.
Let's break that description down piece-by-piece:
**_Chat, messaging, or other natural language interfaces (i.e. voice)
_**The critical element to this new suite of services is that customers can now engage using natural language. It's the difference between opening up the Uber app, navigating the UI, and tapping "Request Pickup" versus simply declaring "I need a ride to this address" in Facebook Messenger.
The conversations can take place in established messaging channels (SMS, WhatsApp, Slack, Telegram, etc), through IP messaging (in-app, web live chat), or even by using voice enabled hardware (Amazon Echo, Siri, etc).
For our purposes, we are not including bots that simply offer a command-line interface to an existing service. For instance, while typing "/giphy party" in Slack does engage with a bot, we don't qualify that as a truly conversational exchange. We realize this is a somewhat subjective guide, but it provides a rough rubric to help us draw the line.
**_Interacting with people, brands, or services and bots
_**The recipient on the other side can be another person, such as a therapist in TalkSpace; a brand, like Sephora on Kik; a service, like AndChill which provides movie recommendations; or a bot, such as the Whole Foods' recipe recommendation engine.
It can be completely staffed by humans, as with Sensay; fully AI, like Poncho's weather bot; or it can take a hybrid approach in using humans + AI tools and suggestions, as is the case with Operator.
**In a bidirectional, asynchronous messaging context**These services provide two-way conversations, both grounded in natural language. This allows for a give-and-take dynamic; for instance, instead of navigating static gift recommendation lists, you can engage in a personal 1-to-1 exchange with a Mezi shopper to narrow your search.
The conversations can be asynchronous ("Amy, please schedule this meeting."), not requiring an immediate answer, or they can take place in real time ("I like them, do you have a pair available in size 12?").
Themes
Segmenting by industry, of course, isn't the only useful way to organize these services. We also sliced them up by how they serve the company: is it a new line of business fundamentally made possible by messaging? An established and "traditional" company adding a messaging product as a supplement? Or perhaps a technology company adding a conversational layer as another way to access its product or service?
New, Messaging-Enabled Companies
The last few years has brought a sudden rise in companies inherently enabled by new messaging technologies. These aren't services utilizing conversation as a form of customer support or an additional contact point; instead, these are companies built entirely around the dynamic of messaging.
Consider Magic or Awesome, text-message assistants that will take care of anything from desk research to travel plans; or more industry-specific services like First Opinion (medical consultation from real doctors) and Riley (lead generation through text message for real estate agents). If you remove the conversational identity from these businesses, you unplug the entire service.
Established Companies
We also looked at established companies dabbling in the promise of conversational commerce. The most basic example is achieved by simply using a messaging channel as a new interface for the core business: order from Domino's by texting them a pizza emoji.
A more interesting and meaningful approach views messaging as fertile ground to try something new. Sephora has built a powerful bot available on Kik that hosts sophisticated interactions with customers. It can gather information about you and your shopping habits, offer tips and reviews about products, and even let you purchase items, all without leaving the messenger.
In this way, Sephora sees their product as a revenue-generating growth opportunity. This approach is not limited to fully-AI chatbots. For instance, Bonobos has embraced live chat to connect shoppers with trained "Ninja" stylists, helping to improve the shopping experience and thus drive sales.
There are myriad instances of traditional companies embracing live chat for customer service. While we could technically make a case to include every business with a substantial social media response team (@CompanyHelp Twitter feeds, for example), we decided to narrow our focus to companies that have made conversational exchanges a significant part of their strategy.
Tech Companies adding a Conversational Channel
In between the two ends of the spectrum are tech-first companies that have added a supplemental messaging product or tie-in. They're native to your phone or desktop, and are working to add a conversational channel in order to achieve greater ubiquity. The messaging interface isn't essential to their business, and is oftentimes just another way to interact with their core offering, but their ethos and technological foundation makes a messaging feature worthwhile and relatively straightforward to implement.
Take Uber and Lyft, for instance, which can be accessed directly within Facebook Messenger after clicking on an address. The service is folded into the messaging interface, but it's output (the ride) is unchanged from the conventional hailing experience.
Compare that to the two-way, natural language experiences found using Hipmunk or Trulia over Facebook Messenger. You can ask for flight/hotel recommendations as if you were engaging with a travel agent, or describe real estate priorities as if talking to a broker.
It's not hard to see other tech firms like Seamless or ZocDoc potentially joining the fray soon: "Get me the usual salad from Y-Cafe," or "See if Doctor Herbert has availability before 10AM any day next week."
Final Thoughts
We put this list together because we wanted to inform our perspective on conversational services, and improve our understanding of companies, bot frameworks, and best-practices within the space.
We're building Argo, a messaging platform for businesses. If you're interested in learning more, please head to ArgoHQ.com or read about our inspiration here.
Finally, if you'd like to make a suggestion or add another service to the list, head to GitHub.com/ArgoHQ/List.